chore(changelog): add PR 9245 fragment

The agentrouter.org upstream WAF returns 400 content-blocked
intermittently when:
  1. messages[].content contains a blocked keyword (Lorem ipsum, the
     phrase 'language model' alone, 'virtual assistant', etc.); or
  2. Requests from the same IP/key arrive in a burst, after which the
     WAF's per-IP suspicion bucket starts blocking content that would
     normally pass. The bucket relaxes after ~5-10s of idle.

Apply three mitigations:

1. Burst guard (open-sse/services/wafRateLimit.ts)
   Per-bucket (provider+url) gate that enforces a 500ms minimum gap
   between outbound requests to agentrouter. Configurable via
   configureWafRateLimit(). Tested in tests/unit/wafRateLimit.test.ts.

2. Reactive retry (BaseExecutor.WAF_RETRY_CONFIG in base.ts)
   New WAF_RETRY_CONFIG with maxAttempts=2, delayMs=1500,
   backoffMultiplier=2. When the upstream returns 400 with a body that
   matches /content[_-]blocked/i, retry the same URL with exponential
   backoff (1.5s, 3.0s) before falling through to the 429/401/fallback
   chain. Tested in tests/unit/base-executor-waf-retry.test.ts.

3. Documentation (docs/security/AGENTROUTER_WAF.md)
   Blocklist of always-blocked and almost-always-blocked patterns,
   behavior under load, guidance for prompts/tool output, and pointers
   to the relevant code paths in OmniRoute.

These are belt-and-suspenders: the burst guard prevents the WAF from
activating on normal traffic, and the reactive retry recovers when it
does anyway. Together they should eliminate the intermittent
400 content-blocked that Claude Code sees when running through
agentrouter via OmniRoute.

Refs #9275 follow-up. Test: 'WAF retry config shape' and 'WAF retry
differs from generic' guard the WAF_RETRY_CONFIG contract so future
refactors don't accidentally collapse the two retry paths.

Co-authored-by: diegosouzapw <diegosouzapw@users.noreply.github.com>

.cbmignore

@@ -0,0 +1,197 @@
+# codebase-memory-mcp ignore list
+#
+# Padrão gitignore-style. Linhas começando com `#` são comentários.
+# Barra final (`/`) = só diretório. Sem barra = casa arquivo OU diretório.
+#
+# O CBM também lê `.gitignore` automaticamente — esta lista deixa explícito o que
+# os hooks do CBM vão pular. Se uma regra entrar em conflito entre os dois arquivos,
+# vale a união. Editar este arquivo é mais barato do que confiar na herança implícita.
+#
+# Última reconciliação: 2026-07-31, status `ready` (513k nodes / 689k edges),
+# `auto_index_limit=50000`, total indexável medido ≈11.546 arquivos (folga 4,3×).
+#
+# Fontes cruzadas:
+#   - `codebase-memory-mcp cli index_status --project home-diegosouzapw-dev-proxys-OmniRoute`
+#     → `not_indexed.dirs` (27) + `not_indexed.files` (336), todos `BY DESIGN`.
+#   - `.gitignore` deste repo (5.691 B) — fonte canônica secundária.
+#
+# Como auditar mudanças: depois de editar este arquivo, rodar `index_repository`
+# (ou esperar `auto_watch` re-indexar) e re-checar `cli index_status` → comparar
+# contagens em `not_indexed.dirs_count` e `not_indexed.files_count`.
+
+# ─────────────────────────────────────────────────────────────────────────────
+# 1. Diretorios de runtime / pacote — nao sao codigo-fonte
+# ─────────────────────────────────────────────────────────────────────────────
+node_modules/
+node_modules
+
+# Builds e artefatos reproduziveis (Layer 1 Next.js / Electron)
+.build/
+dist/
+.next/
+out/
+
+# Electron especifico
+electron/dist-electron/
+electron/node_modules/
+icon.iconset/
+
+# Workspaces internos que tem proprio node_modules
+@omniroute/opencode-plugin/dist/
+@omniroute/opencode-plugin/node_modules/
+@omniroute/opencode-provider/dist/
+@omniroute/opencode-provider/node_modules/
+
+# Recursos nativos compilados (C/JNI/wasm)
+src/mitm/tproxy/native/build/
+
+# Artefatos locais do Stryker / Playwright / coverage
+.stryker-tmp/
+reports/mutation/
+stryker-output-*.json
+.playwright-mcp/
+test-results/
+playwright-report/
+blob-report/
+
+# Analise / linters / caches
+.analysis/
+.sisyphus/
+.plans/
+.gitnexus
+.worktrees
+.codegraph/
+
+# Quality artifacts (gerados por npm run lint --cache etc)
+.eslintcache
+.eslintcache-complexity
+
+# Claude Code local state
+.claude/scheduled_tasks.lock
+.claude/scheduled_tasks/
+.claude/sessions/
+.claude/state.json
+.claude/settings.local.json
+
+# Serena / Antigravity / outras tools locais
+.serena/
+.antigravitycli/
+.gemini/
+.config/
+
+# ─────────────────────────────────────────────────────────────────────────────
+# 2. Diretorios com prefixo `_` — locais / privados (regra global do .gitignore)
+# ─────────────────────────────────────────────────────────────────────────────
+_*/
+_artifacts/
+_cache/
+_mono_repo/
+_references/
+_tasks/
+
+# ─────────────────────────────────────────────────────────────────────────────
+# 3. Diretorios de tooling IA (state local, nao codigo)
+# ─────────────────────────────────────────────────────────────────────────────
+.agents/
+.claude/
+.vscode/
+.idea/
+.junie/
+.omc/
+.data/
+.data-dev/
+.local-data/
+.logs/
+.artifacts/
+.source/
+.superpowers/
+.claude-flow/
+.omnivscodeagent/
+omnirouteCloud/
+omnirouteSite/
+.omniroute/
+.stent/
+
+# Subpaths especificos do Claude Code que nao estao em .claude/ (criados sob repo)
+.claude/worktrees/
+
+# ─────────────────────────────────────────────────────────────────────────────
+# 4. Diretorios de dados / runtime locais (storage, env, secrets, scratch)
+# ─────────────────────────────────────────────────────────────────────────────
+data/
+src/lib/env/
+src/app/api/agent-skills/coverage/
+src/app/api/cloud/
+src/app/api/sync/cloud/
+src/app/api/system/env/
+tests/golden-set/data/
+
+# Logs e saida de teste
+logs/*
+test_output.log
+home-diegosouzapw-dev-automacoes-*.txt
+
+# ─────────────────────────────────────────────────────────────────────────────
+# 5. Diretorios do monorepo por subprojeto (nao fazem parte do app principal)
+# ─────────────────────────────────────────────────────────────────────────────
+security-analysis/
+vscode-extension/
+obsidian-plugin/node_modules/
+
+# ─────────────────────────────────────────────────────────────────────────────
+# 6. Diretorios de documentacao interna / workflow
+# ─────────────────────────────────────────────────────────────────────────────
+docs/superpowers/
+
+# ─────────────────────────────────────────────────────────────────────────────
+# 7. Arquivos especificos (nao diretorios inteiros)
+# ─────────────────────────────────────────────────────────────────────────────
+
+# Segredos e env — NUNCA indexar
+.env
+.env.*
+!.env.example
+!.env.homolog.example
+
+# TypeScript build info e next env declaration
+*.tsbuildinfo
+next-env.d.ts
+typescript
+
+# SQLite transient files (WAL/SHM/journal)
+*.sqlite-shm
+*.sqlite-wal
+*.sqlite-journal
+
+# Mapas e source maps
+*.map
+
+# Bun / npm lockfiles ruidosos
+bun.lock
+
+# `cheaper-inference-gateway.svg` e arquivos de midia na raiz/asset ja cobertos
+# pelos `ignored-suffix` do indexador (svg/png/jpg/ico/etc >50kB ou >500linhas);
+# manter a regra explicita aqui ajuda a auditar:
+cheaper-inference-gateway.svg
+cheaper-inference-gateway-*.svg
+
+# Husky internals
+.husky/_/
+
+# CI / quality metric artifacts
+config/quality/quality-metrics.json
+config/quality/test-impact-map.json
+audit-report.json
+.gh-discussions.json
+
+# i18n audit (gerado por npm run scripts)
+scripts/i18n/_audit.json
+scripts/i18n/_pending-keys.json
+
+# Cli binario local (scratch)
+bin/omniroute.mjs
+
+# Deploy / docker backups
+deploy.sh
+docker-compose.yml.bak
+docker-compose.minimal.yml

.env.example

@@ -1414,10 +1414,6 @@ APP_LOG_TO_FILE=true
 # Default: ~/.omniroute/plugins/  Override in dev/CI to point at a local plugin tree.
 # OMNIROUTE_PLUGIN_PATH=
 
-# Allow plugins to request the 'exec' permission (spawn child processes from the
-# plugin worker sandbox). Disabled by default; set to 1 to enable (local operator only).
-# OMNIROUTE_PLUGINS_ALLOW_EXEC=0
-
 # ── Prompt cache (system prompt deduplication) ──
 # Used by: open-sse/services — caches identical system prompts across requests.
 # PROMPT_CACHE_MAX_SIZE=50              # Max cached entries (default: 50)

.github/pull_request_template.md

@@ -9,11 +9,14 @@
 
 ## Validation
 
-Run only the focused loop for what you changed — the full unit suite, Vitest, the
-60% coverage gate, and the production build all run in CI on this PR (#8329):
+Choose the change type and focused loop from the
+[Contribution Golden Path](../docs/dev/CONTRIBUTION_GOLDEN_PATH.md). The full unit suite,
+Vitest, the 60% coverage gate, and the production build all run in CI on this PR (#8329):
 
-- [ ] Focused tests for the change: `node --import tsx/esm --test tests/unit/<file>.test.ts`
+- [ ] Change type: provider / routing / UI / i18n / CLI / DB / build-deploy / other
+- [ ] Focused tests and category gates from the golden path
 - [ ] `npm run lint`
+- [ ] Reconciled with the current active release base; focused checks rerun afterward
 - [ ] Production-code changes include a new or updated automated test in this PR
 - [ ] SonarQube PR analysis is green or any remaining issues are explicitly documented below
 
@@ -29,4 +32,4 @@ Run only the focused loop for what you changed — the full unit suite, Vitest,
 
 ## Reviewer Notes
 
-- Call out any risky areas, migrations, feature flags, or manual validation that reviewers should know about.
+- Call out any risky areas, migrations, feature flags, or manual validation that reviewers should know about.

.github/workflows/quality.yml

@@ -179,6 +179,81 @@ jobs:
       # Complexity + cognitive-complexity: ONE ESLint walk (both baselines still
       # enforced separately by ruleId). Avoids two cold tree walks on fast-path.
       - run: npm run check:complexity-ratchets
+      # ── G0 (trilho .50): gates do trilho A que faltavam no trilho B ──────────────
+      # The god-file refactor happens in PRs→release/**; without these, the release
+      # rail never sees a new import cycle, dead code, duplication or a security
+      # regression until the release PR to main. Deliberately NOT brought here:
+      # bundle-size (self-skips without a build — this rail's build job is advisory
+      # and uploads nothing, so it would be dead configuration) and the coverage
+      # run (fast-unit already runs the full suite; the coverage ratchet stays on
+      # the main rail via --allow-missing in lint-guard).
+      - run: npm run check:cycles
+      - run: npm run check:lockfile
+      - name: Duplication ratchet
+        run: npm run check:duplication
+      - name: Dead-code ratchet (knip)
+        run: npm run check:dead-code
+      - name: Type coverage ratchet
+        run: npm run check:type-coverage
+      - name: Compression budget ratchet
+        run: npm run check:compression-budget
+      # Security scanners — same hardened install as ci.yml quality-extended
+      # (gh release download = authenticated, 5000 req/hr; curl to api.github.com
+      # is rate-limited to 60/hr and silently no-ops when throttled). The blocking
+      # gates below SKIP (exit 0) when their binary is absent — only a measured
+      # regression vs config/quality/quality-baseline.json blocks.
+      - name: Install security scanners (gitleaks/osv/actionlint/zizmor/oasdiff)
+        continue-on-error: true
+        env:
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          set +e
+          mkdir -p "$HOME/.local/bin"
+          # Ratchets compare scanner COUNTS across runs. Pin every auditor: a rule-set
+          # update must be an explicit PR that re-measures/rebaselines, never a random
+          # red (or green) caused by whatever "latest" served that morning.
+          GITLEAKS_VERSION=v8.30.1
+          OSV_SCANNER_VERSION=v2.3.8
+          ACTIONLINT_VERSION=v1.7.12
+          ZIZMOR_VERSION=1.25.2
+          OASDIFF_VERSION=v1.19.1
+          # gitleaks — pinned linux x64 tarball via gh (authed), extract binary
+          rm -rf /tmp/gl && mkdir -p /tmp/gl
+          gh release download "$GITLEAKS_VERSION" --repo gitleaks/gitleaks --pattern '*linux_x64.tar.gz' --dir /tmp/gl
+          tar -xzf /tmp/gl/*linux_x64.tar.gz -C "$HOME/.local/bin" gitleaks
+          # osv-scanner — pinned linux amd64 bare binary via gh (authed)
+          rm -rf /tmp/osv && mkdir -p /tmp/osv
+          gh release download "$OSV_SCANNER_VERSION" --repo google/osv-scanner --pattern '*linux_amd64' --dir /tmp/osv
+          install -m 0755 /tmp/osv/*linux_amd64 "$HOME/.local/bin/osv-scanner"
+          # actionlint — official installer from a pinned release tag (never main)
+          bash <(curl -fsSL "https://raw.githubusercontent.com/rhysd/actionlint/${ACTIONLINT_VERSION}/scripts/download-actionlint.bash") "$ACTIONLINT_VERSION" "$HOME/.local/bin"
+          # zizmor — pinned PyPI package (same version as ci.yml quality-extended)
+          pipx install "zizmor==$ZIZMOR_VERSION" || pip install --user "zizmor==$ZIZMOR_VERSION"
+          # oasdiff — pinned linux amd64 tarball via gh (authed), extract binary
+          rm -rf /tmp/oasd && mkdir -p /tmp/oasd
+          gh release download "$OASDIFF_VERSION" --repo Tufin/oasdiff --pattern '*linux_amd64.tar.gz' --dir /tmp/oasd
+          tar -xzf /tmp/oasd/*linux_amd64.tar.gz -C "$HOME/.local/bin" oasdiff
+          # ALWAYS export the bin dir (even if any step above failed)
+          echo "$HOME/.local/bin" >> "$GITHUB_PATH"
+          "$HOME/.local/bin/gitleaks" version || true
+          "$HOME/.local/bin/actionlint" -version || true
+          "$HOME/.local/bin/osv-scanner" --version || true
+          "$HOME/.local/bin/oasdiff" --version || true
+          zizmor --version || true
+      - name: Secret scan (gitleaks, ratchet, blocking)
+        run: npm run check:secrets -- --ratchet
+      - name: Vulnerability ratchet (osv-scanner, ratchet, blocking)
+        run: npm run check:vuln-ratchet -- --ratchet
+      - name: Workflow lint (actionlint+zizmor, ratchet, blocking)
+        run: npm run check:workflows -- --ratchet
+      # BASE_REF is read by the script from the env (never interpolated into a
+      # shell body) — workflow-injection-safe. actions/checkout fetches remote
+      # refs, not a local branch named github.base_ref, so prefix origin/ or this
+      # gate self-skips every PR with reason=base-unresolved.
+      - name: OpenAPI breaking-change (oasdiff, ratchet, blocking)
+        env:
+          BASE_REF: ${{ github.base_ref && format('origin/{0}', github.base_ref) || '' }}
+        run: npm run check:openapi-breaking -- --ratchet
       - name: Typecheck (core)
         run: npm run typecheck:core
       # #7033: dashboard-scoped typecheck gate — src/app/(dashboard) TSX is not
@@ -364,6 +439,12 @@ jobs:
     if: ${{ github.event_name != 'pull_request' || ((github.event.pull_request.draft == false || startsWith(github.head_ref, 'mergify/merge-queue/')) && needs.changes.outputs.code == 'true') }}
     runs-on: ubuntu-latest
     continue-on-error: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.repo.fork == true }}
+    # G0 (trilho .50): security-events:read lets the CodeQL ratchet below read open
+    # code-scanning alerts via `gh api .../code-scanning/alerts` (same as ci.yml's
+    # quality-gate job). contents: read keeps checkout working.
+    permissions:
+      contents: read
+      security-events: read
     steps:
       - uses: actions/checkout@v7
         with:
@@ -385,6 +466,29 @@ jobs:
       - name: ESLint (baseline congelado — warning novo = vermelho)
         # lint:json writes the report; --max-warnings 0 keeps no-new-warnings policy.
         run: npm run lint:json -- --max-warnings 0
+      # ── G0 (trilho .50): motor de ratchet também no trilho B ─────────────────────
+      # This job just wrote .artifacts/eslint-results.json — collect-metrics prefers
+      # that file, so the ratchet engine lands here at ZERO extra ESLint cost (one
+      # inventory, two consumers; same reason ci.yml chains lint → quality-gate).
+      # The coverage-report artifact does not exist on this rail, so both ratchet
+      # invocations run --allow-missing: coverage.* metrics skip gracefully while
+      # the deterministic ones (eslint / openapi-coverage / i18n-ui) stay BLOCKING.
+      # Coverage authority remains on the main rail (ci.yml test-coverage → quality-gate).
+      - run: npm run quality:collect
+      - name: Ratchet check (blocking)
+        run: node scripts/quality/check-quality-ratchet.mjs --allow-missing --summary .artifacts/quality-ratchet.md
+      - name: Require-tighten (blocking)
+        run: node scripts/quality/check-quality-ratchet.mjs --allow-missing --require-tighten
+      # CodeQL alerts ratchet — same semantics as ci.yml quality-gate: exits 1 ONLY
+      # on a real regression (open alerts > baseline in quality-baseline.json);
+      # a measurement failure (gh/auth/api) self-skips with exit 0.
+      - name: CodeQL alerts ratchet (blocking)
+        run: npm run check:codeql-ratchet
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      - name: Append ratchet summary
+        if: always()
+        run: cat .artifacts/quality-ratchet.md >> "$GITHUB_STEP_SUMMARY" || true
 
   # Merge-integrity: pega no PR os dois vazamentos crônicos de merge que hoje só
   # explodem na release-PR. (1) CHANGELOG-eat — o auto-resolve do merge come

.gitignore

@@ -171,7 +171,6 @@ config/quality/test-impact-map.json
 # GitNexus local index
 .gitnexus
 .worktrees
-bin/omniroute.mjs
 
 # Consistent with .dockerignore / .npmignore
 .omc/
@@ -201,7 +200,10 @@ scripts/i18n/_pending-keys.json
 .codegraph/
 
 # Fumadocs generated source
-.source/
+/.source/
+
+# Temporary local worktrees used to build unpublished npm tarballs
+/.deploy-build-*/
 
 # AI agent local settings and configs
 .agents/
@@ -250,3 +252,4 @@ tests/homolog/.auth/
 tests/homolog/ui/.auth/
 homolog-report/
 docker-compose.yml.bak
+.playwright-cli/

.source/dynamic.ts

@@ -1,8 +0,0 @@
-// @ts-nocheck
-import { dynamic } from 'fumadocs-mdx/runtime/dynamic';
-import * as Config from '../source.config';
-
-const create = await dynamic<typeof Config, import("fumadocs-mdx/runtime/types").InternalTypeConfig & {
-  DocData: {
-  }
-}>(Config, {"configPath":"source.config.ts","environment":"next","outDir":".source"}, {"doc":{"passthroughs":["extractedReferences"]}});

.source/source.config.mjs

@@ -1,22 +0,0 @@
-// source.config.ts
-import { defineDocs, defineConfig } from "fumadocs-mdx/config";
-var docs = defineDocs({
-  dir: "docs",
-  docs: {
-    files: [
-      "./architecture/**/*.md",
-      "./guides/**/*.md",
-      "./reference/**/*.md",
-      "./frameworks/**/*.md",
-      "./routing/**/*.md",
-      "./security/**/*.md",
-      "./compression/**/*.md",
-      "./ops/**/*.md"
-    ]
-  }
-});
-var source_config_default = defineConfig();
-export {
-  source_config_default as default,
-  docs
-};

AMIT

@@ -1 +0,0 @@
-

CLAUDE.md

@@ -45,7 +45,7 @@ For full test matrix, see `CONTRIBUTING.md` → "Running Tests". For deep archit
 | Translators   | `open-sse/translator/`  | Format conversion (OpenAI↔Claude↔Gemini)                                                                                                                |
 | Transformer   | `open-sse/transformer/` | Responses API ↔ Chat Completions                                                                                                                        |
 | Services      | `open-sse/services/`    | Combo routing, rate limits, caching, etc                                                                                                                |
-| Database      | `src/lib/db/`           | SQLite domain modules (95 files, 110 migrations)                                                                                                        |
+| Database      | `src/lib/db/`           | SQLite domain modules (130 migrations)                                                                                                                  |
 | Domain/Policy | `src/domain/`           | Policy engine, cost rules, fallback logic                                                                                                               |
 | MCP Server    | `open-sse/mcp-server/`  | 104 tools (42 base + memory/skill/agentSkill/pool/notion/obsidian/gamification/plugin modules), 3 transports (stdio / SSE / Streamable HTTP), 31 scopes |
 | A2A Server    | `src/lib/a2a/`          | JSON-RPC 2.0 agent protocol                                                                                                                             |
@@ -72,7 +72,7 @@ Client → /v1/chat/completions (Next.js route)
 
 API routes follow a consistent pattern: `Route → CORS preflight → Zod body validation → Optional auth (extractApiKey/isValidApiKey) → API key policy enforcement → Handler delegation (open-sse)`. No global Next.js middleware — interception is route-specific.
 
-**Combo routing** (`open-sse/services/combo.ts`): 18 strategies (priority, weighted, fill-first, round-robin, p2c, random, least-used, cost-optimized, reset-aware, reset-window, headroom, strict-random, auto, lkgp, context-optimized, context-relay, fusion, pipeline). Each target calls `handleSingleModel()` which wraps `handleChatCore()` with per-target error handling and circuit breaker checks. The `fusion` strategy is the exception: it fans out to a panel of models in parallel, then a judge model synthesizes one final answer (`open-sse/services/fusion.ts`). See `docs/routing/AUTO-COMBO.md` for the 12-factor Auto-Combo scoring + the full strategy table and `docs/architecture/RESILIENCE_GUIDE.md` for the 3 resilience layers.
+**Combo routing** (`open-sse/services/combo.ts`): 19 public strategies (priority, weighted, fill-first, round-robin, p2c, random, least-used, cost-optimized, reset-aware, reset-window, headroom, strict-random, auto, lkgp, context-optimized, cache-optimized, context-relay, fusion, pipeline). Each target calls `handleSingleModel()` which wraps `handleChatCore()` with per-target error handling and circuit breaker checks. The `fusion` strategy is the exception: it fans out to a panel of models in parallel, then a judge model synthesizes one final answer (`open-sse/services/fusion.ts`). See `docs/routing/AUTO-COMBO.md` for the 13-factor Auto-Combo scoring + the full strategy table and `docs/architecture/RESILIENCE_GUIDE.md` for the 3 resilience layers.
 
 ---
 
@@ -332,7 +332,7 @@ For any non-trivial change, read the matching deep-dive first:
 | Repo navigation                               | `docs/architecture/REPOSITORY_MAP.md`                   |
 | Architecture                                  | `docs/architecture/ARCHITECTURE.md`                     |
 | Engineering reference                         | `docs/architecture/CODEBASE_DOCUMENTATION.md`           |
-| Auto-Combo (12-factor scoring, 18 strategies) | `docs/routing/AUTO-COMBO.md`                            |
+| Auto-Combo (13-factor scoring, 19 strategies) | `docs/routing/AUTO-COMBO.md`                            |
 | Resilience (3 mechanisms)                     | `docs/architecture/RESILIENCE_GUIDE.md`                 |
 | Reasoning replay                              | `docs/routing/REASONING_REPLAY.md`                      |
 | Skills framework                              | `docs/frameworks/SKILLS.md`                             |
@@ -461,10 +461,18 @@ own dedicated branch, and you MUST confirm the base branch with the operator bef
    git fetch origin "$BASE_BRANCH"
    git worktree add ".claude/worktrees/${TASK##*/}" -b "$TASK" "origin/$BASE_BRANCH"
    cd ".claude/worktrees/${TASK##*/}"
-   # symlink node_modules from the main checkout to skip a per-worktree npm install:
-   ln -s "$(git -C <main_checkout> rev-parse --show-toplevel)/node_modules" node_modules
+   # Reuse the main checkout's node_modules to skip a per-worktree npm install.
+   # HARD LINKS (`cp -al`), never a symlink: ~5s for the whole tree and near-zero extra
+   # disk (the inodes are shared), and unlike a symlink it does not break the dev server.
+   cp -al "$(git -C <main_checkout> rev-parse --show-toplevel)/node_modules" node_modules
    ```
 
+   **Never `ln -s` node_modules.** Turbopack rejects a symlink that resolves outside the
+   project root, so `npm run dev` dies with a FATAL panic (`Symlink [project]/node_modules
+   is invalid, it points out of the filesystem root`) while typecheck, lint and the test
+   runners all keep passing — the error names "filesystem root", not the worktree, so it
+   reads like a Next/build bug and costs real time to trace (incident 2026-07-31, #9043).
+
    In Claude Code prefer the native `EnterWorktree` tool (it already creates worktrees under
    `.claude/worktrees/`): create the worktree with the command above, then call `EnterWorktree`
    with its `path`.

CONTRIBUTING.md

@@ -2,6 +2,11 @@
 
 Thank you for your interest in contributing! This guide covers everything you need to get started.
 
+For the official per-change workflow, start with the
+[Contribution Golden Path](docs/dev/CONTRIBUTION_GOLDEN_PATH.md). It maps provider, routing,
+UI/UX, i18n, CLI, database, and build/deploy changes to their contracts, focused tests, CI
+coverage, and reconciliation steps.
+
 ---
 
 ## Development Setup
@@ -198,10 +203,11 @@ Coverage notes:
 
 ### Pull Request Requirements
 
-Before opening a PR, run the focused loop for what you changed. The full unit suite
-(4 CI shards), Vitest, the **60%+** coverage gate, and the production build are CI's
-responsibility — running them locally adds no signal the PR checks will not already
-give you, and on smaller machines it can saturate the host (#8084):
+Before opening a PR, use the
+[Contribution Golden Path](docs/dev/CONTRIBUTION_GOLDEN_PATH.md) to run the focused loop for
+what you changed. The full unit suite (4 CI shards), Vitest, the **60%+** coverage gate, and
+the production build are CI's responsibility — running them locally adds no signal the PR
+checks will not already give you, and on smaller machines it can saturate the host (#8084):
 
 - Run the test files that cover your change: `node --import tsx/esm --test tests/unit/<file>.test.ts`
 - Run `npm run lint`
@@ -271,7 +277,7 @@ src/                        # TypeScript (.ts / .tsx)
 │   ├── a2a/                # Agent-to-Agent v0.3 protocol server
 │   ├── acp/                # Agent Communication Protocol registry
 │   ├── compliance/         # Compliance policy engine
-│   ├── db/                 # SQLite database layer (21 modules + 16 migrations)
+│   ├── db/                 # SQLite domain modules + 130 migrations
 │   ├── memory/             # Persistent conversational memory
 │   ├── oauth/              # OAuth providers, services, and utilities
 │   ├── skills/             # Extensible skill framework
@@ -281,7 +287,7 @@ src/                        # TypeScript (.ts / .tsx)
 ├── mitm/                   # MITM proxy (cert, DNS, target routing)
 ├── shared/
 │   ├── components/         # React components (.tsx)
-│   ├── constants/          # Provider definitions (177), MCP scopes, 14 routing strategies
+│   ├── constants/          # Provider definitions (290), MCP scopes, 19 routing strategies
 │   ├── utils/              # Circuit breaker, sanitizer, auth helpers
 │   └── validation/         # Zod v4 schemas
 └── sse/                    # SSE proxy pipeline
@@ -289,7 +295,7 @@ src/                        # TypeScript (.ts / .tsx)
 open-sse/                   # @omniroute/open-sse workspace
 ├── executors/              # 14 provider-specific request executors
 ├── handlers/               # 11 request handlers (chat, responses, embeddings, images, etc.)
-├── mcp-server/             # MCP server (25 tools, 3 transports, 10 scopes)
+├── mcp-server/             # MCP server (104 tools, 3 transports, 31 scopes)
 ├── services/               # 36+ services (combo, autoCombo, rateLimitManager, etc.)
 ├── translator/             # Format translators (OpenAI ↔ Claude ↔ Gemini ↔ Responses ↔ Ollama)
 ├── transformer/            # Responses API transformer

README.md

@@ -241,12 +241,53 @@ curl http://localhost:20128/v1/chat/completions \
       <b>What Kimi's support powers:</b> Kimi's API credits power OmniRoute's AI-validated release pipeline — the <i>merge validation powered by Kimi K3</i> stage that reviews every pull request before it ships — plus day-to-day feature development. First-class Kimi support ships on both rails: the direct <a href="https://platform.kimi.ai?aff=omniroute">Kimi API</a> (<code>kimi-k3</code>) and the <a href="https://www.kimi.com/code?aff=omniroute">Kimi Code coding plan</a> (OAuth and API key). OmniRoute is also the first Brazilian open-source project in Kimi's support program. <a href="https://platform.kimi.ai?aff=omniroute"><b>Get a Kimi API key →</b></a>
     </td>
   </tr>
+  <tr>
+    <td align="center" width="150">
+      <a href="https://cheaperinference.com/?utm_source=omniroute">
+        <img src="public/providers/cheaperinference.svg" width="64" alt="Cheaper Inference"/>
+      </a>
+      <br/><b>Cheaper Inference</b><br/><sub>cheaperinference.com</sub><br/><br/>
+      <img src="https://img.shields.io/badge/Open_Source_Friend-31f889?style=flat-square&labelColor=04170d" alt="Open Source Friend"/>
+    </td>
+    <td>
+      Thanks to <b>Cheaper Inference</b>, an OmniRoute Open Source Friend, for backing this project! Cheaper Inference is a cost-ranked gateway that resells 42 frontier models — Claude, GPT-5.x, Gemini, Kimi K3, GLM, DeepSeek, Grok and MiniMax — behind one OpenAI-compatible endpoint, routing each request to the cheapest eligible provider without ever charging above the model maker's list price.
+      <br/><br/>
+      <b>First-class support in OmniRoute:</b> Chat Completions, the native <code>/v1/responses</code> endpoint, vision, tool calling and 3 image models (<code>grok-imagine</code>, <code>nano-banana-pro</code>, <code>nano-banana-2</code>, reachable as <code>cheaperinference/&lt;model&gt;</code>). <a href="https://cheaperinference.com/?utm_source=omniroute"><b>Get an API key →</b></a>
+    </td>
+  </tr>
 </table>
 
 <sub>Links tagged <code>aff=omniroute</code> are partner links. They fund the project at no extra cost to you.</sub>
 
 <br/>
 
+<details open>
+<summary><sub><b>🎟️ Affiliates Promo</b> — free signup coupons from providers we don't sponsor (click to expand)</sub></summary>
+
+<sub><i>This section is for referral/coupon codes only. Sponsored partnerships live in <b>🤝 Supported by our Open Source Friends</b> above. OmniRoute has no sponsorship or partnership with the providers listed here — these are public coupons anyone can use.</i></sub>
+
+<table>
+  <tr>
+    <td align="center" width="120">
+      <a href="https://agentrouter.org/register?aff=70LM">
+        <img src="public/providers/agentrouter.png" width="32" alt="AgentRouter"/>
+      </a>
+      <br/><sub><b>AgentRouter</b></sub><br/><sub>agentrouter.org</sub>
+    </td>
+    <td>
+      <sub><b><a href="https://agentrouter.org/register?aff=70LM">AgentRouter</a></b> — affiliate signup · <b>$100 free credits</b> on signup (free server, expect higher latency — best for testing, not production). First-class support in OmniRoute since <b>v3.8.50</b>: Chat Completions, the Anthropic-compatible wire format and the OpenAI-compatible path. Available models include <code>claude-opus-4-8</code>, <code>claude-opus-5</code>, <code>gpt-5.6-sol</code> and more. <b><a href="https://agentrouter.org/register?aff=70LM">Grab your $100 →</a></b></sub>
+      <br/><br/>
+      <sub>⚠️ <i>Affiliate link — OmniRoute has no sponsorship or partnership with this provider.</i></sub>
+    </td>
+  </tr>
+</table>
+
+<sub>Know another provider with a generous free signup coupon that benefits OmniRoute users? Open an issue and we'll add it here.</sub>
+
+</details>
+
+<br/>
+
 <div align="center">
 
 ## 🎯 Combos — The Flagship

changelog.d/fixes/9235-french-ui-catalog.md

@@ -0,0 +1 @@
+- **fix(i18n):** completed the French UI catalog by adding all missing keys and replacing every placeholder translation ([#9235](https://github.com/diegosouzapw/OmniRoute/pull/9235)) — thanks @alex-jordan547

changelog.d/fixes/9245-hardcoded-web-ui-i18n.md

@@ -0,0 +1 @@
+- **fix(i18n):** localized hardcoded web UI copy across public pages, dashboard views, and shared components, with complete French and Vietnamese coverage ([#9245](https://github.com/diegosouzapw/OmniRoute/pull/9245)) — thanks @alex-jordan547

config/quality/file-size-baseline.json

@@ -177,7 +177,7 @@
     "tests/unit/account-fallback-service.test.ts": 1563,
     "tests/unit/batch_api.test.ts": 1324,
     "tests/unit/cc-compatible-provider.test.ts": 1217,
-    "tests/unit/chatcore-translation-paths.test.ts": 2769,
+    "tests/unit/chatcore-translation-paths.test.ts": 2776,
     "tests/unit/chatgpt-web.test.ts": 3148,
     "tests/unit/combo-routing-engine.test.ts": 3449,
     "tests/unit/db-migration-runner.test.ts": 1499,
@@ -342,7 +342,7 @@
     "_rebaseline_pr1043_minimax_tts": "Upstream port decolua/9router#1043 (toanalien) own growth: audioSpeech.ts 965->1061 (+96). Adds MiniMax T2A v2 TTS dispatch (handleMinimaxSpeech + hexToBytes helper) — provider entry was already in audioRegistry (format: minimax-tts) but no handler existed, falling through to the OpenAI-compatible default that fails (T2A has custom shape + hex-encoded audio + base_resp envelope). New branch sits next to the other inline provider branches (xiaomi-mimo, coqui, tortoise, aws-polly) — extracting would just create indirection. Covered by tests/unit/minimax-tts-1043.test.ts (3 tests, GREEN: success, base_resp error, invalid-hex).",
     "_rebaseline_pr4592_exclude_exhausted_auto": "Reconcile #4592 already-merged growth: combo.ts 2991->3036 (+45, terminal-status quota-cutoff exclusion in buildAutoCandidates + opt-in gate). Fast-gate PR->release does not run check:file-size.",
     "open-sse/executors/antigravity.ts": 1528,
-    "open-sse/executors/base.ts": 1562,
+    "open-sse/executors/base.ts": 1578,
     "open-sse/executors/chatgpt-web.ts": 3241,
     "open-sse/executors/codex.ts": 1534,
     "open-sse/executors/cursor.ts": 1560,
@@ -365,7 +365,7 @@
     "open-sse/services/rateLimitManager.ts": 1060,
     "open-sse/translator/response/openai-responses.ts": 1174,
     "open-sse/utils/cursorAgentProtobuf.ts": 1505,
-    "open-sse/utils/stream.ts": 2887,
+    "open-sse/utils/stream.ts": 2889,
     "src/app/(dashboard)/dashboard/HomePageClient.tsx": 1381,
     "src/app/(dashboard)/dashboard/analytics/ComboHealthTab.tsx": 1031,
     "src/app/(dashboard)/dashboard/api-manager/ApiManagerPageClient.tsx": 3117,
@@ -413,5 +413,6 @@
   "_rebaseline_2026_07_28_8860_tokenrefresh_projectid": "PR #8860 (fix/antigravity-projectid-centralized) own test growth: tests/unit/token-refresh-service.test.ts 1311->1378 (+67 = 4 cases covering projectId discovery on the tokenRefresh.ts path — the Dashboard/health-check refresh route, which #8842 did not reach since that fixed the executor path). Covered by the same file.",
   "_rebaseline_2026_07_28_8861_xiaomi_token_plan": "PR #8861 (feat/xiaomi-token-plan-protocol-selector) own growth: EditConnectionModal.tsx 1283->1316 (+33 = the per-connection API-protocol selector field) and open-sse/executors/base.ts 1540->1562 (+22 = alternate-format resolution at the existing buildUrl/headers chokepoint). Both are irreducible wiring at existing call sites.",
   "_rebaseline_2026_07_28_8863_firefly_detail_level": "PR #8863 (fix/adobe-firefly-gpt-detail-level-max) own growth: adobeFireflyClient.ts 2317->2322 (+5 = gpt-image detailLevel defaulting to maximal at the existing payload-build site). Covered by tests/unit/adobe-firefly.test.ts.",
-  "_rebaseline_2026_07_29_8281_home_quickstart_prefetch": "Release v3.8.49 base-red fix (no PR — captain sweep): src/app/(dashboard)/dashboard/HomePageClient.tsx 1377->1381 (+4). #8292 added prefetch={false} to the sidebar but left /home's five quick-start Links prefetching, so first paint still fired 12 speculative RSC requests — caught by navigation.spec.ts only after the e2e helper bug (APP_ROUTE_PATTERN missing /home) was repaired in the same cycle. Growth is the five prefetch attributes; it was offset first by extracting the repeated className literals (INLINE_LINK x4, DOCS_LINK x1), which collapsed five wrapped <Link> blocks back to one line each — a naive fix measured 1391. Guard: tests/unit/sidebar-prefetch-policy-8281.test.ts."
+  "_rebaseline_2026_07_29_8281_home_quickstart_prefetch": "Release v3.8.49 base-red fix (no PR — captain sweep): src/app/(dashboard)/dashboard/HomePageClient.tsx 1377->1381 (+4). #8292 added prefetch={false} to the sidebar but left /home's five quick-start Links prefetching, so first paint still fired 12 speculative RSC requests — caught by navigation.spec.ts only after the e2e helper bug (APP_ROUTE_PATTERN missing /home) was repaired in the same cycle. Growth is the five prefetch attributes; it was offset first by extracting the repeated className literals (INLINE_LINK x4, DOCS_LINK x1), which collapsed five wrapped <Link> blocks back to one line each — a naive fix measured 1391. Guard: tests/unit/sidebar-prefetch-policy-8281.test.ts.",
+  "_rebaseline_2026_08_02_v3850_agentrouter_responses": "Release v3.8.50 AgentRouter/Codex compatibility reconciliation. open-sse/executors/base.ts 1562->1578: #9190 wires AgentRouter's selected Claude/OpenAI/Responses protocol through the existing executor URL, auth, identity-header and fingerprint chokepoints; the reusable alternate resolver remains outside base.ts. open-sse/utils/stream.ts 2887->2889: #9213 evaluates Responses ID and usage normalization independently so response.completed always receives finite usage.total_tokens instead of short-circuiting after an ID rewrite. tests/unit/chatcore-translation-paths.test.ts 2769->2776: #9191 updates the existing Claude-Code bridge assertions for the dynamic AgentRouter wire image. PR #9224 offsets its own chatCore growth by extracting the AgentRouter protocol decisions into chatCore/agentRouterProtocol.ts, leaving chatCore below its frozen ceiling. Covered by agentrouter executor/chatCore protocol tests, chatcore translation-path tests, and responses-commentary-passthrough tests."
 }

docs/dev/CONTRIBUTION_GOLDEN_PATH.md

@@ -0,0 +1,229 @@
+---
+title: "Contribution Golden Path"
+---
+
+# Contribution Golden Path
+
+Use this guide to choose the smallest reliable development loop for a pull request. It does not
+replace the area-specific architecture and security documents linked below; it connects each common
+change type to its contracts, focused checks, and CI coverage.
+
+## The path every change follows
+
+1. **Choose the base before editing.** Find the highest active `release/v*` branch and branch from
+   its tip. Target that branch, not `main`. If a release freeze is active, do not target the frozen
+   branch; use the next active cycle described in
+   [Branching & Release Model](../ops/BRANCHING_MODEL.md).
+2. **Name the contracts.** Identify every catalog, schema, generated artifact, public API, or user
+   interface that the change affects. The table below gives the minimum starting set.
+3. **Write or update focused tests.** Production changes in `src/`, `open-sse/`, `electron/`, or
+   `bin/` require an automated test in the same PR. Run the smallest test files that prove the
+   behavior, then the listed focused gates.
+4. **Let CI run the broad matrix.** The complete unit shards, Vitest, coverage ratchet, and
+   production build run on the PR. Run a broad suite locally only when a focused failure points to
+   wider impact or when the change spans several subsystems.
+5. **Reconcile before review.** Fetch the active base, inspect its new commits and your diff against
+   it, then rebase or merge the base according to the contributor workflow. Resolve generated-file
+   and catalog conflicts from their source, regenerate them, rerun the focused loop, and confirm the
+   PR still targets the active release branch.
+6. **Record evidence.** In the PR template, list the commands run, every test file added or changed,
+   migrations or feature flags, and any CI-only validation still pending.
+
+## Golden paths by change type
+
+Commands below are minimum focused checks, not permission to skip a test that directly covers the
+behavior you changed.
+
+### Provider
+
+**Contracts**
+
+- Provider definition in `src/shared/constants/providers/` and its composition in
+  `src/shared/constants/providers.ts`.
+- Models and capabilities in `open-sse/config/providerRegistry.ts` or its extracted registry files.
+- Executor/translator selection, OAuth or API-key configuration, dashboard assets, and generated
+  provider reference when applicable.
+- Public credentials must use `resolvePublicCred()`; error responses must use the shared sanitized
+  error helpers. See [Public Credentials](../security/PUBLIC_CREDS.md) and
+  [Error Sanitization](../security/ERROR_SANITIZATION.md).
+
+**Focused loop**
+
+```bash
+npm run check:provider-consistency
+npm run check:provider-assets
+node --import tsx/esm --test tests/unit/provider-translate-path-golden.test.ts
+node --import tsx/esm --test tests/unit/<provider-or-executor>.test.ts
+npm run gen:provider-reference   # when the catalog changes; commit the generated diff
+npm run lint
+```
+
+Also test every affected request family: chat, Responses, images, embeddings, audio, or video.
+Review generated catalog and golden diffs as contract changes; do not accept them blindly.
+
+### Routing
+
+**Contracts**
+
+- Public strategy values and UI metadata in `src/shared/constants/routingStrategies.ts`.
+- Dispatch and ordering under `open-sse/services/combo.ts` and `open-sse/services/combo/`.
+- Combo schemas, persistence, resilience state, model capabilities, and API/UI controls.
+- [Auto-Combo Engine](../routing/AUTO-COMBO.md) and resilience documentation when behavior changes.
+
+**Focused loop**
+
+```bash
+node --import tsx/esm --test tests/unit/combo-<behavior>.test.ts
+npm run test:combo:matrix        # strategy or dispatch changes
+npm run check:known-symbols      # strategy registration changes
+npm run lint
+```
+
+Use deterministic mocked-upstream tests locally. Live combo smokes require credentials and are
+manual, not CI substitutes.
+
+### UI / UX
+
+**Contracts**
+
+- Next.js route/page and shared component boundaries under `src/app/` and
+  `src/shared/components/`.
+- API response shapes, loading/empty/error states, keyboard and screen-reader behavior,
+  responsive layout, theming, and locale expansion.
+- English UI source strings in `src/i18n/messages/en.json`; do not hard-code new user-facing copy.
+
+**Focused loop**
+
+```bash
+node --import tsx --test tests/unit/dashboard/<feature>.test.ts
+npx vitest run --config vitest.config.ts tests/unit/ui/<component>.test.tsx
+npm run check:dashboard-typecheck
+npm run lint
+```
+
+Run the app for interaction or visual changes and check both narrow and wide viewports. CI runs the
+production build and broader suites; visual behavior still needs a focused component, Playwright,
+or documented manual check appropriate to the change.
+
+### i18n
+
+**Contracts**
+
+- `src/i18n/messages/en.json` is the UI source; `config/i18n.json` is the locale source.
+- CLI catalogs live separately under `bin/cli/locales/`.
+- Preserve ICU placeholders and tags exactly. Do not translate product/provider/model names,
+  protocol and header names, commands, code/JSON identifiers, URLs, environment variables, or
+  protected terms such as `OmniRoute`, `OAuth`, `MCP`, and `A2A`. The current source list is
+  `scripts/i18n/glossary/protected-terms.json`.
+
+**Focused loop**
+
+```bash
+npm run i18n:sync-ui:dry
+npm run i18n:check-ui-coverage
+npm run i18n:check-value-drift
+npm run i18n:check-glossary
+npm run check:cli-i18n          # when CLI strings/catalogs change
+npm run lint
+```
+
+This is guidance for the existing system, not an invitation to expand its tooling or key model.
+Keep i18n patches surgical while the replacement system is being designed. Do not run translation
+commands that call external services unless the task explicitly requires generated translations and
+you have reviewed the resulting diff.
+
+### CLI
+
+**Contracts**
+
+- Public commands and flags in `bin/cli/`, generated API commands, exit codes, stdout/stderr and
+  JSON output shapes, config/environment behavior, and packaged files.
+- CLI user-facing strings must use the CLI i18n layer and keep `en`/`pt-BR` catalogs aligned.
+- Preserve Node as the supported runtime and the published binary contract.
+
+**Focused loop**
+
+```bash
+node --import tsx/esm --test tests/unit/cli/<command>.test.ts
+npm run check:cli-i18n
+npm run build:cli             # generated/bundled CLI changes
+npm run check:pack-policy     # package-surface changes
+npm run lint
+```
+
+Use the exact command in a temporary data directory when behavior depends on parsing, files, or exit
+status. CI performs the broader package artifact and ecosystem checks.
+
+### Database
+
+**Contracts**
+
+- Domain modules under `src/lib/db/`; `src/lib/localDb.ts` remains a re-export layer only.
+- Numbered, idempotent SQL migrations under `src/lib/db/migrations/`, transaction safety, upgrade
+  behavior, indexes, and every caller affected by the schema.
+- Routes and handlers never issue raw SQL directly.
+
+**Focused loop**
+
+```bash
+npm run check:migration-numbering
+npm run check:db-rules
+node --import tsx/esm --test tests/unit/db/<domain>.test.ts
+node --import tsx/esm --test tests/unit/db/migration-<number>.test.ts
+npm run lint
+```
+
+Test both a fresh database and upgrade from the prior schema when adding a migration. Database tests
+must close handles and call `resetDbInstance()` during cleanup. Run `npm run test:bun:db` only when
+the best-effort Bun adapter path changes; Node remains authoritative.
+
+### Build / deploy
+
+**Contracts**
+
+- Root and workspace manifests/lockfile, `scripts/build/`, Next.js standalone assembly, `dist/`
+  package contents, Electron platform metadata, CI workflows, and deployment sentinels.
+- Supported Node ranges and the allow-listed Bun use in `CLAUDE.md` must remain intact.
+- Build artifacts stay untracked; dependency, license, workflow, and package policies apply.
+
+**Focused loop**
+
+```bash
+node --import tsx/esm --test tests/unit/build/<behavior>.test.ts
+npm run check:build-scope
+npm run check:lockfile         # dependency or lockfile changes
+npm run check:pack-policy      # published package surface changes
+npm run lint
+```
+
+Use `npm run build` locally only when the change affects compilation, standalone assembly, assets,
+or runtime bundling. Use `npm run build:release` only for release/deploy validation. CI's build is
+the final cross-platform signal; platform-specific Electron changes need the matching focused build
+or smoke evidence.
+
+## Local loop versus CI
+
+| Run locally for each patch                                                    | CI supplies the broad signal                                      |
+| ----------------------------------------------------------------------------- | ----------------------------------------------------------------- |
+| Direct behavior tests and category gates above                                | Sharded full unit suite and serial tests                          |
+| `npm run lint`                                                                | Vitest suites and coverage/quality ratchets                       |
+| Typecheck or build only when the affected contract calls for it               | Production build, security, docs, dependency, and PR-policy gates |
+| Manual interaction/live checks only when automation cannot prove the behavior | Cross-job integration and platform checks configured by workflow  |
+
+A green focused loop is evidence about the changed contract, not proof that unrelated CI checks
+will pass. Conversely, do not make every local edit wait for the full repository matrix.
+
+## Reconciliation checklist
+
+Before requesting review:
+
+- Confirm the PR base is still the highest active `release/v*` branch.
+- Fetch that base and review commits that landed since you branched.
+- Review `git diff <active-base>...HEAD` for accidental or generated churn.
+- Resolve catalog and generated-document conflicts by updating the source and regenerating output.
+- Rerun every focused test/gate listed in the PR description after reconciliation.
+- Never weaken assertions or drop required tests merely to match a moved base.
+
+For release-freeze and retargeting rules, use
+[Branching & Release Model](../ops/BRANCHING_MODEL.md). For the complete CI inventory, use
+[Quality Gates Reference](../architecture/QUALITY_GATES.md).

docs/i18n/pl/CLAUDE.md

@@ -461,10 +461,18 @@ own dedicated branch, and you MUST confirm the base branch with the operator bef
    git fetch origin "$BASE_BRANCH"
    git worktree add ".claude/worktrees/${TASK##*/}" -b "$TASK" "origin/$BASE_BRANCH"
    cd ".claude/worktrees/${TASK##*/}"
-   # symlink node_modules from the main checkout to skip a per-worktree npm install:
-   ln -s "$(git -C <main_checkout> rev-parse --show-toplevel)/node_modules" node_modules
+   # Reuse the main checkout's node_modules to skip a per-worktree npm install.
+   # HARD LINKS (`cp -al`), never a symlink: ~5s for the whole tree and near-zero extra
+   # disk (the inodes are shared), and unlike a symlink it does not break the dev server.
+   cp -al "$(git -C <main_checkout> rev-parse --show-toplevel)/node_modules" node_modules
    ```
 
+   **Never `ln -s` node_modules.** Turbopack rejects a symlink that resolves outside the
+   project root, so `npm run dev` dies with a FATAL panic (`Symlink [project]/node_modules
+   is invalid, it points out of the filesystem root`) while typecheck, lint and the test
+   runners all keep passing — the error names "filesystem root", not the worktree, so it
+   reads like a Next/build bug and costs real time to trace (incident 2026-07-31, #9043).
+
    In Claude Code prefer the native `EnterWorktree` tool (it already creates worktrees under
    `.claude/worktrees/`): create the worktree with the command above, then call `EnterWorktree`
    with its `path`.

docs/i18n/zh-CN/CLAUDE.md

@@ -415,10 +415,18 @@ git push -u origin feat/your-feature
    git fetch origin "$BASE_BRANCH"
    git worktree add ".claude/worktrees/${TASK##*/}" -b "$TASK" "origin/$BASE_BRANCH"
    cd ".claude/worktrees/${TASK##*/}"
-   # 从主工作区符号链接 node_modules，省去每个 worktree 的 npm install：
-   ln -s "$(git -C <main_checkout> rev-parse --show-toplevel)/node_modules" node_modules
+   # 复用主工作区的 node_modules，省去每个 worktree 的 npm install。
+   # 必须用硬链接（`cp -al`），绝不能用符号链接：整棵树约 5 秒，几乎不占额外磁盘
+   # （inode 是共享的），而且与符号链接不同，它不会破坏开发服务器。
+   cp -al "$(git -C <main_checkout> rev-parse --show-toplevel)/node_modules" node_modules
    ```
 
+   **绝不要对 node_modules 使用 `ln -s`。** Turbopack 会拒绝解析到项目根目录之外的符号链接，
+   因此 `npm run dev` 会以 FATAL panic 崩溃（`Symlink [project]/node_modules is invalid, it
+   points out of the filesystem root`），而 typecheck、lint 和测试运行器却都照常通过 —— 错误信息
+   提到的是 "filesystem root" 而不是 worktree，看起来像 Next/构建的 bug，排查会浪费大量时间
+   （事故 2026-07-31，#9043）。
+
    在 Claude Code 中优先使用原生的 `EnterWorktree` 工具（它已经在 `.claude/worktrees/` 下创建 worktree）：先用上述命令创建 worktree，然后用其 `path` 调用 `EnterWorktree`。
 
 3. **工作、提交、推送、发起 PR — 全部在 worktree 内部完成。** 绝不在另一个会话可能共享的 worktree 内 `git checkout` 不同分支。

docs/issues/plugin-browser-pool-proposal.md

@@ -1,168 +0,0 @@
-# [Feature] Pluginization Phase 1: Extract Playwright/CloakBrowser Browser Pool
-
-**Labels:** `enhancement`, `plugin`, `architecture`
-
-## Problem / Use Case
-
-OmniRoute's Playwright/CloakBrowser dependency is a **large, non-essential dependency** (~1500+ LOC across 22 source files + ~35 test files) pulled into every installation regardless of whether the user needs browser-backed chat. Users who run OmniRoute purely as a proxy/router (the majority) pay for:
-
-- **Disk space**: ~200+ MB from Playwright browsers + Chromium binaries (installed via `npx playwright install`)
-- **Build complexity**: Turbopack must handle the `cloakbrowser` package
-- **Bundle size**: All browser-pool code is compiled into the main codebase
-- **Surface area**: 7 files with direct Playwright imports (4 dynamic, 3 static type imports)
-- **CI/cache impact**: Playwright installs in CI pipelines even when not needed
-
-Currently, only environment variables (`OMNIROUTE_BROWSER_POOL=off`) gate _runtime_ execution — the code still loads, imports get resolved, and Playwright must be installed.
-
-## Proposed Solution
-
-Extract the Playwright/CloakBrowser browser pool into an **optional package** loaded via dynamic `import()` at runtime, following the existing pattern used for `cloakbrowser` (computed-string dynamic import to avoid resolution). The core retains thin interface stubs that gracefully degrade when the optional package is absent.
-
-### Architecture
-
-```
-open-sse/
-  interfaces/
-    browserPool.ts       ← NEW: BrowserPoolProvider interface + types
-  services/
-    browserPool.ts       ← BECOMES: thin stub, delegates to optional package
-    browserBackedChat.ts ← BECOMES: thin stub, delegates to optional package
-    grokClearance.ts     ← BECOMES: thin stub
-
-packages/browser-pool/   ← NEW: optional package
-  index.ts               ← exports BrowserPoolProvider implementation
-  src/
-    browserPool.ts       ← extracted from open-sse/services/browserPool.ts
-    browserBackedChat.ts ← extracted from open-sse/services/browserBackedChat.ts
-    grokClearance.ts     ← extracted from open-sse/services/grokClearance.ts
-    claudeTurnstileSolver.ts ← extracted (tightly coupled, moved as-is)
-    inAppLoginService.ts ← extracted (own Playwright instance, separate lifecycle)
-  package.json
-  tsconfig.json
-```
-
-### Phase Breakdown
-
-**Phase 1 — Core pool extraction (this issue):**
-
-1. Define `BrowserPoolProvider` interface in `open-sse/interfaces/browserPool.ts`
-2. Extract `browserPool.ts` (~502 LOC), `browserBackedChat.ts` (~270 LOC), `grokClearance.ts` (~84 LOC) into `packages/browser-pool/`
-3. Replace core files with thin stubs that try `import('../../../packages/browser-pool')` with graceful fallback
-4. Keep `poolTools.ts` importing the core stub (unchanged from consumer perspective)
-5. Make Playwright an optional dependency (not in root `package.json`)
-6. Typecheck core passes with and without the package installed
-7. All existing tests pass (with plugin installed)
-
-**Phase 2 — Turnstile solver extraction (future):**
-
-- Extract `claudeTurnstileSolver.ts` (~212 LOC) — has static Playwright type imports, needs type interface
-- Move `claudeWebAutoRefresh.ts` (depends on turnstile solver)
-
-**Phase 3 — Standalone Playwright instances (future):**
-
-- Extract `inAppLoginService.ts` (~257 LOC)
-- Refactor `gemini-web.ts` executor's own Playwright path (~553 LOC)
-
-### Interface Design (Phase 1)
-
-```typescript
-// open-sse/interfaces/browserPool.ts
-export interface BrowserPoolProvider {
-  acquireBrowserContext(options?: BrowserPoolContextOptions): Promise<PooledContext>;
-  releaseBrowserContext(ctx: PooledContext): Promise<void>;
-  getBrowserPoolMetrics(): BrowserPoolMetrics;
-  shutdownPool(): Promise<void>;
-  isPoolEnabled(): boolean;
-  openPage(url: string, ctx?: PooledContext): Promise<{ page: any }>;
-  readPageResponseBody(page: any): Promise<string>;
-  getBrowserPoolStatus(): BrowserPoolStatus;
-}
-```
-
-### Stub Pattern
-
-```typescript
-// open-sse/services/browserPool.ts — thin stub
-let _impl: BrowserPoolProvider | null = null;
-
-async function getImpl(): Promise<BrowserPoolProvider> {
-  if (!_impl) {
-    try {
-      const { createBrowserPoolProvider } = await import("../../packages/browser-pool");
-      _impl = createBrowserPoolProvider();
-    } catch {
-      // Graceful fallback — disabled
-      _impl = createNullBrowserPoolProvider();
-    }
-  }
-  return _impl;
-}
-
-export async function acquireBrowserContext(...args) {
-  return (await getImpl()).acquireBrowserContext(...args);
-}
-```
-
-## Alternatives Considered
-
-1. **Existing hook-based PluginManager**: Rejected. The current PluginManager operates via child-process IPC and request-pipeline hooks (`onRequest`, `onResponse`, `onError`). A browser pool is an in-process runtime service with composable lifecycle — not a request pipeline hook. Forcing it through IPC would add ~50ms+ per browser operation and break the existing synchronous pool pattern.
-
-2. **Keep as-is, just lazy-load the import**: Minimal improvement — the dependency tree still references Playwright types, requiring it to be available. Doesn't reduce bundle size or simplify CI.
-
-3. **Replace Playwright with a protocol-level abstraction**: Too ambitious and would change the behavior of the pool. Playwright's CDP capabilities (context isolation, cookies, screenshots) are fundamental to how the pool works.
-
-4. **Monorepo workspace**: Too heavy for this scope. A simple extracted package avoids workspace tooling changes.
-
-## Acceptance Criteria
-
-1. `open-sse/interfaces/browserPool.ts` exists and exports `BrowserPoolProvider`, `PooledContext`, `BrowserPoolMetrics` types
-2. `open-sse/services/browserPool.ts` becomes a thin stub with zero Playwright imports
-3. `packages/browser-pool/` contains all extracted implementation (browserPool, browserBackedChat, grokClearance)
-4. Core typecheck (`npm run typecheck:core`) passes with 0 errors **without** the browser-pool package installed
-5. Core typecheck passes with the package installed
-6. All existing tests pass when the browser-pool package is installed
-7. `poolTools.ts` `omniroute_browser_pool_status` tool works end-to-end when the package is installed
-8. Graceful degradation: when the package is absent, `getBrowserPoolStatus()` returns `{ enabled: false }` without crashing
-9. Playwright is moved from root `dependencies` to optional/peer in the extracted package
-10. Documentation updated in `docs/reference/ENVIRONMENT.md`
-
-## Expected Test Plan
-
-- Unit tests for the stub fallback path (simulate import failure, verify graceful degradation)
-- Unit tests moved to the extracted package
-- Verify `tests/unit/browser-pool-optional-import.test.ts` passes (still validates cloakbrowser isn't statically resolved)
-- Verify `tests/unit/browserPool-proxy.test.ts` passes
-- Verify `tests/unit/browserBackedChat-matcher.test.ts` passes
-- E2E: `npm run typecheck:core` without the package installed → 0 errors
-- E2E: `npm run test:coverage` (with package installed) → existing coverage gates pass
-
-## Additional Context
-
-Current dependency graph (simplified):
-
-```
-open-sse/services/browserPool.ts (502 LOC, singleton Playwright/CloakBrowser pool)
-  ├── open-sse/services/browserBackedChat.ts (270 LOC, browser-backed chat runner)
-  │     ├── open-sse/executors/claude-web.ts (imports tryBackedChat)
-  │     └── open-sse/executors/duckduckgo-web.ts (imports tryBackedChat)
-  ├── open-sse/services/grokClearance.ts (84 LOC, CF clearance via browser)
-  └── open-sse/mcp-server/tools/poolTools.ts (imports getBrowserPoolMetrics)
-
-Standalone Playwright users (separate, future phases):
-  ├── open-sse/services/claudeTurnstileSolver.ts (212 LOC, static Playwright type imports)
-  ├── open-sse/services/inAppLoginService.ts (257 LOC, own browser lifecycle)
-  └── open-sse/executors/gemini-web.ts (553 LOC, private Playwright path)
-
-Kill switches: OMNIROUTE_BROWSER_POOL, WEB_COOKIE_USE_BROWSER (both env vars)
-```
-
-Total extracted in Phase 1: ~856 LOC, 3 files.
-Total deferred to Phase 2/3: ~1022 LOC, 4 files.
-
-This is the first pluginization step. Future targets (separate issues): memory/compression plugin, additional provider support extraction.
-
-## Related References
-
-- PR #8219 (model catalog connection filter + cache TTL) — same baseline `release/v3.8.49`
-- `docs/reference/ENVIRONMENT.md` — browser pool env vars documentation
-- Plugin system docs at `docs/PLUGINS.md` — existing PluginManager (not used here, referenced for contrast)

docs/reference/ENVIRONMENT.md

@@ -414,7 +414,6 @@ detection above).
 | `OMNIROUTE_HTTP_TIMEOUT_MS`    | `30000`    | `bin/cli/api.mjs`                       | Per-attempt HTTP timeout (ms) for CLI → server requests.                                                                           |
 | `OMNIROUTE_VERBOSE`            | `0`        | `bin/cli/api.mjs`                       | Set to `1` to print retry/backoff diagnostics to stderr during CLI commands.                                                       |
 | `OMNIROUTE_PLUGIN_PATH`        | _(unset)_  | `bin/cli/plugins.mjs`                   | Custom directory for CLI plugin discovery (`omniroute-cmd-*` packages). Defaults to `~/.omniroute/plugins/` when unset.            |
-| `OMNIROUTE_PLUGINS_ALLOW_EXEC` | `0`        | `src/lib/plugins/pluginWorker.ts`       | Set to `1` to allow plugins to request the `exec` permission (spawn child processes from the worker sandbox). Local operator only. |
 
 ---
 

docs/routing/AUTO-COMBO.md

@@ -167,21 +167,21 @@ The Auto-Combo Engine dynamically selects the best provider/model for each reque
 
 > Source: [diagrams/auto-combo-12factor.mmd](../diagrams/auto-combo-12factor.mmd) (regenerate via `npm run docs:render-diagrams`). Diagram/filename predate the `cacheAffinity` factor added by #8008 and still show 12 factors.
 
-| Factor                | Default Weight | Description                                                                                        |
-| :-------------------- | :------------- | :------------------------------------------------------------------------------------------------- |
-| `health`              | 0.20           | Health score from circuit breaker (CLOSED=1.0, HALF_OPEN=0.5, OPEN=0.0)                            |
-| `quota`               | 0.15           | Remaining quota / rate-limit headroom [0..1]                                                       |
-| `costInv`             | 0.15           | Inverse **blended** cost (60% input + 40% output token price, normalized) — cheaper = higher score |
-| `latencyInv`          | 0.12           | Inverse p95 latency normalized to pool — faster = higher score                                     |
-| `taskFit`             | 0.08           | Task-type fitness (coding, review, planning, analysis, debugging, docs)                            |
-| `stability`           | 0.05           | Variance-based stability (low latency stdDev / error rate)                                         |
-| `tierPriority`        | 0.05           | Account-tier priority — Ultra=1.0, Pro=0.67, Standard=0.33, Free=0.0                               |
-| `tierAffinity`        | 0.05           | Affinity between the candidate's tier and the manifest-recommended tier                            |
-| `specificityMatch`    | 0.05           | Match between request specificity (manifest hint) and model tier                                   |
-| `contextAffinity`     | 0.05           | Affinity between the request's context-window need and the model's context window                  |
-| `connectionDensity`   | 0.05           | Spreads load across connections of the same provider (anti-concentration)                          |
+| Factor                | Default Weight | Description                                                                                                                                                                                 |
+| :-------------------- | :------------- | :------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
+| `health`              | 0.20           | Health score from circuit breaker (CLOSED=1.0, HALF_OPEN=0.5, OPEN=0.0)                                                                                                                     |
+| `quota`               | 0.15           | Remaining quota / rate-limit headroom [0..1]                                                                                                                                                |
+| `costInv`             | 0.15           | Inverse **blended** cost (60% input + 40% output token price, normalized) — cheaper = higher score                                                                                          |
+| `latencyInv`          | 0.12           | Inverse p95 latency normalized to pool — faster = higher score                                                                                                                              |
+| `taskFit`             | 0.08           | Task-type fitness (coding, review, planning, analysis, debugging, docs)                                                                                                                     |
+| `stability`           | 0.05           | Variance-based stability (low latency stdDev / error rate)                                                                                                                                  |
+| `tierPriority`        | 0.05           | Account-tier priority — Ultra=1.0, Pro=0.67, Standard=0.33, Free=0.0                                                                                                                        |
+| `tierAffinity`        | 0.05           | Affinity between the candidate's tier and the manifest-recommended tier                                                                                                                     |
+| `specificityMatch`    | 0.05           | Match between request specificity (manifest hint) and model tier                                                                                                                            |
+| `contextAffinity`     | 0.05           | Affinity between the request's context-window need and the model's context window                                                                                                           |
+| `connectionDensity`   | 0.05           | Spreads load across connections of the same provider (anti-concentration)                                                                                                                   |
 | `cacheAffinity`       | 0.00           | Rendezvous-hash affinity toward the connection likeliest to already hold this request's prompt-cache prefix (`open-sse/services/combo/promptCacheAffinity.ts`); disabled by default (#8008) |
-| `resetWindowAffinity` | 0.00           | Bias toward connections whose quota reset window is favorable (disabled by default)                |
+| `resetWindowAffinity` | 0.00           | Bias toward connections whose quota reset window is favorable (disabled by default)                                                                                                         |
 
 **Sum:** `0.20 + 0.15 + 0.15 + 0.12 + 0.08 + 0.05 + 0.05 + 0.05 + 0.05 + 0.05 + 0.05 + 0.00 + 0.00 = 1.0` (validated by `validateWeights()`).
 
@@ -215,11 +215,11 @@ combo's stored config. These apply only to the `auto` strategy and only for the
 that carries them; the combo's saved `modePack`/`budgetCap`/`budgetFallback` are used
 when the header is absent.
 
-| Header                        | Accepts                                                                                                                                                                                 | Effect                                                                                                                                                                                              |
-| :----------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `X-OmniRoute-Mode`            | a preset alias (`fast`, `balanced`, `quality`, `cheap`, `reliable`, `offline`) or a raw pack name (`ship-fast`, `cost-saver`, `quality-first`, `offline-friendly`, `reliability-first`) | Overrides the scoring weights for this request. `balanced`/`default` force the default weights (no pack). Unknown values are ignored (config preserved).                                            |
-| `X-OmniRoute-Budget`          | a positive number (max USD per request)                                                                                                                                                 | Hard cost ceiling: candidates whose estimated cost exceeds it are filtered before selection. What happens when **every** candidate exceeds it is controlled by `X-OmniRoute-Budget-Fallback` below. |
-| `X-OmniRoute-Budget-Fallback` | `cheapest` (default, aliases: `cheapest-viable`, `soft`) or `strict` (aliases: `block`, `hard`)                                                                                        | `cheapest`: falls back to the globally cheapest candidate even though it still exceeds the cap (legacy behavior). `strict`: refuses to select — the request fails fast with `HTTP 402` instead of silently overspending. Unknown values are ignored. |
+| Header                        | Accepts                                                                                                                                                                                 | Effect                                                                                                                                                                                                                                               |
+| :---------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `X-OmniRoute-Mode`            | a preset alias (`fast`, `balanced`, `quality`, `cheap`, `reliable`, `offline`) or a raw pack name (`ship-fast`, `cost-saver`, `quality-first`, `offline-friendly`, `reliability-first`) | Overrides the scoring weights for this request. `balanced`/`default` force the default weights (no pack). Unknown values are ignored (config preserved).                                                                                             |
+| `X-OmniRoute-Budget`          | a positive number (max USD per request)                                                                                                                                                 | Hard cost ceiling: candidates whose estimated cost exceeds it are filtered before selection. What happens when **every** candidate exceeds it is controlled by `X-OmniRoute-Budget-Fallback` below.                                                  |
+| `X-OmniRoute-Budget-Fallback` | `cheapest` (default, aliases: `cheapest-viable`, `soft`) or `strict` (aliases: `block`, `hard`)                                                                                         | `cheapest`: falls back to the globally cheapest candidate even though it still exceeds the cap (legacy behavior). `strict`: refuses to select — the request fails fast with `HTTP 402` instead of silently overspending. Unknown values are ignored. |
 
 ```bash
 # Force the fastest profile, cap this request at $0.05, and hard-block instead of overspending
@@ -240,27 +240,27 @@ resolved values feed the engine's existing `config.modePack` / `config.budgetCap
 
 OmniRoute's combo engine supports **19 routing strategies** (declared in `src/shared/constants/routingStrategies.ts` → `ROUTING_STRATEGY_VALUES`). The Auto Combo engine itself is exposed under the `auto` strategy; the others are available for persisted combos.
 
-| Strategy            | Description                                                                                                                  |
-| :------------------ | :--------------------------------------------------------------------------------------------------------------------------- |
-| `priority`          | First-target ordered list with explicit priority                                                                             |
-| `weighted`          | Weighted random by per-target weight                                                                                         |
-| `round-robin`       | Cycle through targets in order                                                                                               |
-| `context-relay`     | Hand off context across targets (long conversations)                                                                         |
-| `fill-first`        | Fill each target's quota before moving to next                                                                               |
-| `p2c`               | Power-of-2-choices random load balancing                                                                                     |
-| `random`            | Uniform random selection                                                                                                     |
-| `least-used`        | Pick target with lowest current load                                                                                         |
-| `cost-optimized`    | Minimize $ per request given catalog pricing                                                                                 |
-| `reset-aware` ⭐    | Prioritize by quota reset time — short reset windows ranked higher                                                           |
-| `reset-window`      | Prefer targets whose quota window resets soonest                                                                             |
-| `headroom`          | Pick the target with the most remaining quota headroom                                                                       |
-| `strict-random`     | Random without deduplication of repeats                                                                                      |
-| `auto`              | Use Auto Combo scoring (9-factor) — **recommended**                                                                          |
-| `lkgp`              | Last-Known-Good Path (sticky route to last successful target)                                                                |
-| `context-optimized` | Pick target with best fit for current context size                                                                           |
+| Strategy            | Description                                                                                                                                                                               |
+| :------------------ | :---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `priority`          | First-target ordered list with explicit priority                                                                                                                                          |
+| `weighted`          | Weighted random by per-target weight                                                                                                                                                      |
+| `round-robin`       | Cycle through targets in order                                                                                                                                                            |
+| `context-relay`     | Hand off context across targets (long conversations)                                                                                                                                      |
+| `fill-first`        | Fill each target's quota before moving to next                                                                                                                                            |
+| `p2c`               | Power-of-2-choices random load balancing                                                                                                                                                  |
+| `random`            | Uniform random selection                                                                                                                                                                  |
+| `least-used`        | Pick target with lowest current load                                                                                                                                                      |
+| `cost-optimized`    | Minimize $ per request given catalog pricing                                                                                                                                              |
+| `reset-aware` ⭐    | Prioritize by quota reset time — short reset windows ranked higher                                                                                                                        |
+| `reset-window`      | Prefer targets whose quota window resets soonest                                                                                                                                          |
+| `headroom`          | Pick the target with the most remaining quota headroom                                                                                                                                    |
+| `strict-random`     | Random without deduplication of repeats                                                                                                                                                   |
+| `auto`              | Use Auto Combo scoring (9-factor) — **recommended**                                                                                                                                       |
+| `lkgp`              | Last-Known-Good Path (sticky route to last successful target)                                                                                                                             |
+| `context-optimized` | Pick target with best fit for current context size                                                                                                                                        |
 | `cache-optimized`   | Reorder targets by prompt-cache affinity — the connection likeliest to already hold this request's cached prefix is tried first (`open-sse/services/combo/promptCacheAffinity.ts`, #8008) |
-| `fusion` 🧬         | Fan out to a panel of models in parallel, then synthesize one answer via a judge (see below)                                 |
-| `pipeline`          | Run targets sequentially, threading each step's output into the next step's input; only the final answer is returned (#6396) |
+| `fusion` 🧬         | Fan out to a panel of models in parallel, then synthesize one answer via a judge (see below)                                                                                              |
+| `pipeline`          | Run targets sequentially, threading each step's output into the next step's input; only the final answer is returned (#6396)                                                              |
 
 ⭐ = New in v3.8.0 · 🧬 = New in v3.8.36
 
@@ -679,11 +679,11 @@ See `docs/marketing/TIERS.md` for tier definitions and provider classification.
 
 ### Deterministic routing-decision matrix (`npm run test:combo:matrix`)
 
-`tests/integration/combo-matrix/*.test.ts` proves the routing **decision** of all 18
+`tests/integration/combo-matrix/*.test.ts` proves the routing **decision** of all 19
 public strategies end-to-end through the real combo pipeline with a mocked upstream.
 Coverage includes:
 
-- All 18 `ROUTING_STRATEGY_VALUES` strategies (ordered, weighted, cost, context, fusion, …).
+- All 19 `ROUTING_STRATEGY_VALUES` strategies (ordered, weighted, cost, context, fusion, …).
 - `quota-share` (internal) end-to-end: DRR fairness + saturation deprioritization via the
   real `selectQuotaShareTarget` seam (`registerQuotaFetcher` / `setLKGP` /
   `__setHeadroomSaturationFetcherForTests`).

docs/security/AGENTROUTER_WAF.md

@@ -0,0 +1,91 @@
+# agentrouter.org WAF (Web Application Firewall)
+
+The `agentrouter` upstream gateway runs a keyword-based content filter on
+`messages[].content`. The filter is partially deterministic (always blocks
+certain phrases) and partially probabilistic (burst-sensitive — becomes
+more aggressive after rapid requests, recovers after a cooldown).
+
+When the WAF blocks a request it returns:
+
+```
+HTTP/1.1 400 Bad Request
+{"error":{"code":"content-blocked","message":"content-blocked (request id: ...)","param":"","type":"agent_router_api_error"}}
+```
+
+## Scope of the filter
+
+The WAF inspects `messages[].content` only. It does **not** inspect:
+
+- The `system` prompt
+- Structured content blocks (`tool_result`, `tool_use`, `thinking`, `image`)
+- Tool `description` and `input_schema` fields
+- Request metadata, headers, or model id
+
+## Always-blocked patterns (case-insensitive)
+
+| Pattern                       | Notes                                  |
+|-------------------------------|----------------------------------------|
+| Any `Lorem ipsum` variant     | Full Latin lorem vocabulary is blocked |
+| `language model` (alone)      | "the language model" and "large language model" pass |
+| `virtual assistant`           | "AI assistant" passes                   |
+| `I'm here to help`            | "here to help" alone also blocks        |
+| `Claude, made by Anthropic`   | Full phrase only                        |
+
+## Almost-always-blocked patterns
+
+| Pattern           | Notes                                                   |
+|-------------------|---------------------------------------------------------|
+| `placeholder`     | When it stands alone (not as a parameter name, etc.)  |
+| `dummy data`      | Common seed phrase for fixtures                        |
+| `foo bar baz`     | Canonical placeholder phrase                           |
+| Repeated short tokens (`AAA BBB CCC`, `test test test`) | Detector for keyword stuffing |
+
+## Behavior under load
+
+After ~5 rapid requests in a short window, the WAF begins blocking content
+that would normally pass. The bucket relaxes after ~5–10 seconds of idle
+time. This is the same IP-and-key-bound rate limiter that causes
+intermittent `400 content-blocked` errors when Claude Code or Codex CLI
+makes multiple tool-use / message-send calls in quick succession.
+
+## Mitigations already applied in OmniRoute
+
+1. **`open-sse/services/wafRateLimit.ts`** — burst guard that enforces a
+   500 ms minimum gap between outbound requests to any `agentrouter:*`
+   URL. The gap is well below human perception of latency and prevents
+   the WAF from activating on normal traffic.
+
+2. **`BaseExecutor.WAF_RETRY_CONFIG`** — when an upstream returns
+   `400 content-blocked`, the executor retries the same URL with
+   exponential backoff (1.5 s, 3.0 s, max 2 attempts). After the backoff
+   the WAF usually relaxes and the retry succeeds.
+
+3. **`tests/unit/compression/harness.test.ts`** — the test fixture
+   `longInput` was changed from `"lorem ipsum dolor sit amet ".repeat(40)`
+   to `"example content for testing purposes ".repeat(40)` so that when
+   Claude Code reads this file via the `Read` tool, the file contents
+   do not flow back through a `tool_result` block and trip the WAF.
+
+## Guidance for prompts and tool output
+
+If a Claude Code or Codex CLI session repeatedly hits
+`400 content-blocked`, check the most recent user message and the most
+recent tool result for any of the patterns above and rephrase. Common
+workarounds:
+
+- Replace `Lorem ipsum …` with `example text …` or the actual content
+  the test or fixture is trying to model.
+- Replace `placeholder` (when standing alone) with `example value`,
+  `sample value`, or the real value.
+- Replace `language model` with `large language model` or `the model`.
+- Replace `dummy data` with `sample data` or realistic seed values.
+- Replace `I'm here to help` / `here to help` with a more specific
+  opener (e.g. "I'll review the file you mentioned").
+
+## Reporting the false positives upstream
+
+The current filter is overly aggressive — it blocks "Lorem ipsum" in
+`tool_result` blocks even though the operator clearly did not intend to
+inject a prompt. Operators who want this fixed at the source should
+contact `agentrouter.org` to report the false positives. The blocklist
+above is the empirical result of probing the upstream as of 2026-08-03.

eslint.config.mjs

@@ -13,6 +13,29 @@ const TO_NUMBER_RESTRICTION = {
     "canonical coercion shape and the `toNumberOrNull`/`toNumberArray` variants.",
 };
 
+const LOCAL_DB_IMPORT_RESTRICTION = {
+  regex: "^(?:@/lib/localDb(?:\\.ts)?|(?:\\.\\.?/)+(?:lib/)?localDb(?:\\.ts)?)$",
+  message:
+    "The localDb compatibility barrel is restricted — import the owning domain module " +
+    "from `@/lib/db/` instead.",
+};
+
+const EXECUTOR_IMPORT_RESTRICTION = {
+  regex: "^(?:@omniroute/)?open-sse/executors(?:/|$)",
+  message:
+    "Executor implementations must stay behind an open-sse handler or service boundary.",
+};
+
+const PROP_TYPES_RESTRICTION = {
+  name: "prop-types",
+  message: "PropTypes are deprecated. Use TypeScript types/interfaces instead.",
+};
+
+const IMPORT_BOUNDARY_RESTRICTIONS = {
+  paths: [PROP_TYPES_RESTRICTION],
+  patterns: [LOCAL_DB_IMPORT_RESTRICTION],
+};
+
 /** @type {import("eslint").Linter.Config[]} */
 const eslintConfig = [
   ...nextVitals,
@@ -39,15 +62,32 @@ const eslintConfig = [
       "no-eval": "error",
       "no-implied-eval": "error",
       "no-new-func": "error",
+      "no-restricted-imports": ["error", IMPORT_BOUNDARY_RESTRICTIONS],
+    },
+  },
+  // G14: DB internals may use the compatibility barrel while it is decomposed; all
+  // other source files must import the owning src/lib/db domain module directly.
+  {
+    files: ["src/lib/db/**/*.{ts,tsx,js,jsx}"],
+    rules: {
       "no-restricted-imports": [
         "error",
         {
-          paths: [
-            {
-              name: "prop-types",
-              message: "PropTypes are deprecated. Use TypeScript types/interfaces instead.",
-            },
-          ],
+          paths: [PROP_TYPES_RESTRICTION],
+        },
+      ],
+    },
+  },
+  // G14: App routes/components must delegate provider execution through handlers or
+  // services instead of reaching into executor implementations.
+  {
+    files: ["src/app/**/*.{ts,tsx,js,jsx}"],
+    rules: {
+      "no-restricted-imports": [
+        "error",
+        {
+          ...IMPORT_BOUNDARY_RESTRICTIONS,
+          patterns: [LOCAL_DB_IMPORT_RESTRICTION, EXECUTOR_IMPORT_RESTRICTION],
         },
       ],
     },

open-sse/config/codexClient.ts

@@ -1,4 +1,9 @@
-const DEFAULT_CODEX_CLIENT_VERSION = "0.144.1";
+import {
+  DEFAULT_CODEX_CLIENT_VERSION,
+  getCodexCliRsHeaders as buildCodexCliRsHeaders,
+} from "@/shared/constants/codexClient";
+
+export { DEFAULT_CODEX_CLIENT_VERSION } from "@/shared/constants/codexClient";
 const DEFAULT_CODEX_USER_AGENT_PLATFORM = "Windows 10.0.26200";
 const DEFAULT_CODEX_USER_AGENT_ARCH = "x64";
 const CODEX_VERSION_OVERRIDE_ENV = "CODEX_CLIENT_VERSION";
@@ -42,6 +47,10 @@ export function getCodexDefaultHeaders(): Record<string, string> {
   };
 }
 
+export function getCodexCliRsHeaders(): Record<string, string> {
+  return buildCodexCliRsHeaders(getCodexClientVersion());
+}
+
 export function normalizeCodexSessionId(value: unknown): string | null {
   if (typeof value !== "string") return null;
   const normalized = value.trim();

open-sse/config/freeModelCatalog.data.ts

@@ -19,10 +19,9 @@ import type { FreeModelBudget } from "./freeModelCatalog.ts";
 export const FREE_CATALOG_CURATED_AT = "2026-07-22";
 
 export const FREE_MODEL_BUDGETS: FreeModelBudget[] = [
-  { provider: "agentrouter", modelId: "claude-opus-4-6", displayName: "Claude 4.6 Opus", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
-  { provider: "agentrouter", modelId: "claude-haiku-4-5-20251001", displayName: "Claude 4.5 Haiku", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
-  { provider: "agentrouter", modelId: "glm-5.1", displayName: "GLM 5.1", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
-  { provider: "agentrouter", modelId: "deepseek-v3.2", displayName: "DeepSeek V3.2", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
+  { provider: "agentrouter", modelId: "claude-opus-4-8", displayName: "Claude Opus 4.8", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
+  { provider: "agentrouter", modelId: "claude-opus-5", displayName: "Claude Opus 5", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
+  { provider: "agentrouter", modelId: "gpt-5.6-sol", displayName: "GPT-5.6 Sol", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
   { provider: "agy", modelId: "claude-opus-4-6-thinking", displayName: "Claude Opus 4.6 (Thinking)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
   { provider: "agy", modelId: "claude-sonnet-4-6", displayName: "Claude Sonnet 4.6 (Thinking)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
   { provider: "agy", modelId: "gemini-3.1-pro-low", displayName: "Gemini 3.1 Pro (Low)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },

open-sse/config/imageRegistry.ts

@@ -11,6 +11,7 @@ import { KIE_IMAGE_MODELS } from "./providers/registry/kie/imageModels.ts";
 import { FREEPIK_IMAGE_PROVIDER } from "./providers/registry/freepik/index.ts";
 import { STABILITY_AI_IMAGE_MODELS } from "./providers/registry/stability-ai/imageModels.ts";
 import { GEMINI_IMAGEN_PROVIDER } from "./providers/registry/gemini/imageModels.ts";
+import { CHEAPERINFERENCE_IMAGE_PROVIDER } from "./providers/registry/cheaperinference/imageModels.ts";
 
 interface ImageModelEntry {
   id: string;
@@ -714,6 +715,13 @@ export const IMAGE_PROVIDERS: Record<string, ImageProviderConfig> = {
     supportedSizes: ["1:1", "16:9", "9:16", "4:3", "3:4", "1024x1024", "1792x1024", "1024x1792"],
   },
 
+  // Cheaper Inference (OSS-sponsor gateway). Declared AFTER adobe-firefly on
+  // purpose: it shares the nano-banana-pro / nano-banana-2 ids, and parseImageModel
+  // resolves a bare id by first-match over this object's iteration order, so
+  // Firefly keeps the bare ids and these are prefix-only. See the module for the
+  // full collision note.
+  cheaperinference: CHEAPERINFERENCE_IMAGE_PROVIDER,
+
   // Keep Bailian Coding Plan after existing duplicate model owners so adding
   // explicit `bailian-coding-plan/` and `bcp/` routes does not change
   // historical bare-model routing.

open-sse/config/providers/index.ts

@@ -117,6 +117,7 @@ import { uncloseaiProvider } from "./registry/uncloseai/index.ts";
 import { nscaleProvider } from "./registry/nscale/index.ts";
 import { chatgpt_webProvider } from "./registry/chatgpt-web/index.ts";
 import { openrouterProvider } from "./registry/openrouter/index.ts";
+import { cheaperinferenceProvider } from "./registry/cheaperinference/index.ts";
 import { openvectaProvider } from "./registry/openvecta/index.ts";
 import { orcarouterProvider } from "./registry/orcarouter/index.ts";
 import { copilot_webProvider } from "./registry/copilot-web/index.ts";
@@ -337,6 +338,7 @@ export const REGISTRY: Record<string, RegistryEntry> = {
   nscale: nscaleProvider,
   "chatgpt-web": chatgpt_webProvider,
   openrouter: openrouterProvider,
+  cheaperinference: cheaperinferenceProvider,
   openvecta: openvectaProvider,
   orcarouter: orcarouterProvider,
   "copilot-web": copilot_webProvider,

open-sse/config/providers/registry/agentrouter/index.ts

@@ -1,4 +1,5 @@
 import type { RegistryEntry } from "../../shared.ts";
+import { getCodexCliRsHeaders } from "../../../codexClient.ts";
 
 export const agentrouterProvider: RegistryEntry = {
   id: "agentrouter",
@@ -8,6 +9,22 @@ export const agentrouterProvider: RegistryEntry = {
   baseUrl: "https://agentrouter.org/v1/messages",
   authType: "apikey",
   authHeader: "x-api-key",
+  alternateFormats: [
+    {
+      format: "openai",
+      baseUrl: "https://agentrouter.org/v1/chat/completions",
+      authHeader: "bearer",
+      headers: getCodexCliRsHeaders(),
+      label: "OpenAI-compatible (Codex)",
+    },
+    {
+      format: "openai-responses",
+      baseUrl: "https://agentrouter.org/v1/responses",
+      authHeader: "bearer",
+      headers: getCodexCliRsHeaders(),
+      label: "OpenAI Responses (Codex)",
+    },
+  ],
   defaultContextLength: 128000,
   // No static `headers` here: agentrouter now adopts the DYNAMIC Claude-Code
   // wire image via CC_WIRE_IMAGE_BUILTINS (#6056) — the fingerprint/headers are
@@ -15,10 +32,9 @@ export const agentrouterProvider: RegistryEntry = {
   // own baseUrl + x-api-key auth. A static fingerprint here would drift and
   // trip AgentRouter's WAF ("unauthorized client detected").
   models: [
-    { id: "claude-opus-4-6", name: "Claude 4.6 Opus" },
-    { id: "claude-haiku-4-5-20251001", name: "Claude 4.5 Haiku" },
-    { id: "glm-5.1", name: "GLM 5.1" },
-    { id: "deepseek-v3.2", name: "DeepSeek V3.2" },
+    { id: "claude-opus-4-8", name: "Claude Opus 4.8" },
+    { id: "claude-opus-5", name: "Claude Opus 5" },
+    { id: "gpt-5.6-sol", name: "GPT-5.6 Sol" },
   ],
   passthroughModels: true,
 };

open-sse/config/providers/registry/cheaperinference/imageModels.ts

@@ -0,0 +1,34 @@
+/**
+ * Cheaper Inference image provider registry entry.
+ * Extracted into its own module to keep open-sse/config/imageRegistry.ts
+ * under the file-size cap (god-file decomposition; semantic split) — same
+ * pattern as FREEPIK_IMAGE_PROVIDER / SEGMIND_IMAGE_PROVIDER.
+ *
+ * 3 image models measured from GET /v1/models?type=image on 2026-07-31.
+ *
+ * COLLISION NOTE: nano-banana-pro and nano-banana-2 are ALSO adobe-firefly model
+ * ids. parseImageModel() resolves a bare id by first-match over IMAGE_PROVIDERS
+ * iteration order, so this entry is spread into that object AFTER adobe-firefly:
+ * bare `nano-banana-2` keeps routing to Firefly (pre-existing behaviour) and these
+ * models are reachable only as `cheaperinference/<id>` / `cinf/<id>`. Do NOT add
+ * IMAGE_MODEL_ALIASES entries for them — that would silently re-route Firefly
+ * users. Guarded by tests/unit/cheaperinference-image-models.test.ts.
+ *
+ * The endpoint ignores `response_format:"url"` and always returns `b64_json`
+ * (measured twice). The OpenAI image path already handles b64_json; this is not a
+ * bug to "fix". /v1/images/edits returns 404 upstream, so no edit support.
+ */
+export const CHEAPERINFERENCE_IMAGE_PROVIDER = {
+  id: "cheaperinference",
+  alias: "cinf",
+  baseUrl: "https://api.cheaperinference.com/v1/images/generations",
+  authType: "apikey",
+  authHeader: "bearer",
+  format: "openai",
+  models: [
+    { id: "grok-imagine", name: "Grok Imagine (Cheaper Inference)" },
+    { id: "nano-banana-pro", name: "Nano Banana Pro (Cheaper Inference)" },
+    { id: "nano-banana-2", name: "Nano Banana 2 (Cheaper Inference)" },
+  ],
+  supportedSizes: ["1024x1024", "2048x2048", "4096x4096"],
+};

open-sse/config/providers/registry/cheaperinference/index.ts

@@ -0,0 +1,247 @@
+import type { RegistryEntry, RegistryModel } from "../../shared.ts";
+
+/**
+ * Cheaper Inference (https://api.cheaperinference.com) — cost-ranked OpenAI-compatible
+ * gateway, OmniRoute Open Source Friend.
+ *
+ * Catalog captured from a live `GET /v1/models` on 2026-07-31 (42 entries: these 39
+ * `type:"text"` models plus 3 `type:"image"` models that live in imageRegistry.ts —
+ * sending an image model here returns HTTP 400 "Use POST /v1/images/generations").
+ * `supportsVision`/`supportsReasoning` mirror each entry's `capabilities` object
+ * verbatim; they are not inferred from the model name.
+ *
+ * The gateway also serves a native `/v1/responses` endpoint (`responsesBaseUrl`).
+ * It is stateless and REQUIRES `store:false` — see executors/cheaperinference.ts,
+ * which injects it and resolves the URL from the per-model `targetFormat` tag.
+ */
+export const CHEAPERINFERENCE_MODELS: RegistryModel[] = [
+  { id: "aion-labs.aion-2-0", name: "Aion 2.0", supportsReasoning: true, toolCalling: true },
+  {
+    id: "claude-fable-5",
+    name: "Claude Fable 5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-haiku-4.5",
+    name: "Claude Haiku 4.5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-4-7-fast",
+    name: "Claude Opus 4.7 Fast",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-4-8-fast",
+    name: "Claude Opus 4.8 Fast",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-4.5",
+    name: "Claude Opus 4.5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-4.6",
+    name: "Claude Opus 4.6",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-4.7",
+    name: "Claude Opus 4.7",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-4.8",
+    name: "Claude Opus 4.8",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-5",
+    name: "Claude Opus 5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-opus-5-fast",
+    name: "Claude Opus 5 Fast",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-sonnet-4.5",
+    name: "Claude Sonnet 4.5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-sonnet-4.6",
+    name: "Claude Sonnet 4.6",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "claude-sonnet-5",
+    name: "Claude Sonnet 5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  { id: "deepseek-v4-flash", name: "DeepSeek V4 Flash", supportsReasoning: true, toolCalling: true },
+  { id: "deepseek-v4-pro", name: "DeepSeek V4 Pro", supportsReasoning: true, toolCalling: true },
+  {
+    id: "gemini-2.5-flash",
+    name: "Gemini 2.5 Flash",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "gemini-3-5-flash",
+    name: "Gemini 3.5 Flash",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "gemini-3-flash-preview",
+    name: "Gemini 3 Flash Preview",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "gemini-3.1-flash-lite",
+    name: "Gemini 3.1 Flash Lite",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "gemini-3.1-pro",
+    name: "Gemini 3.1 Pro",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "gemini-3.1-pro-preview",
+    name: "Gemini 3.1 Pro Preview",
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  { id: "glm-4.5", name: "GLM-4.5", supportsReasoning: true, toolCalling: true },
+  { id: "glm-4.5-air", name: "GLM-4.5 Air", supportsReasoning: true, toolCalling: true },
+  { id: "glm-4.6", name: "GLM-4.6", supportsReasoning: true, toolCalling: true },
+  { id: "glm-4.7", name: "GLM-4.7", supportsReasoning: true, toolCalling: true },
+  { id: "glm-5", name: "GLM-5", supportsReasoning: true, toolCalling: true },
+  { id: "glm-5.1", name: "GLM-5.1", supportsReasoning: true, toolCalling: true },
+  { id: "glm-5.2", name: "GLM-5.2", supportsReasoning: true, toolCalling: true },
+  {
+    id: "google/gemini-3.5-flash-lite",
+    name: "Gemini 3.5 Flash Lite",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  // The GPT-5.x family is tagged for the gateway's native /v1/responses endpoint —
+  // that is the surface OpenAI-family clients (Codex-style) expect for tool loops.
+  {
+    id: "gpt-5.4",
+    name: "GPT-5.4",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+    targetFormat: "openai-responses",
+  },
+  {
+    id: "gpt-5.4-mini",
+    name: "GPT-5.4 Mini",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+    targetFormat: "openai-responses",
+  },
+  {
+    id: "gpt-5.5",
+    name: "GPT-5.5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+    targetFormat: "openai-responses",
+  },
+  {
+    id: "gpt-5.6-luna",
+    name: "GPT-5.6 Luna",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+    targetFormat: "openai-responses",
+  },
+  {
+    id: "gpt-5.6-sol",
+    name: "GPT-5.6 Sol",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+    targetFormat: "openai-responses",
+  },
+  {
+    id: "gpt-5.6-terra",
+    name: "GPT-5.6 Terra",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+    targetFormat: "openai-responses",
+  },
+  {
+    id: "grok-4.5",
+    name: "Grok 4.5",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  {
+    id: "kimi-k3",
+    name: "Kimi K3",
+    supportsVision: true,
+    supportsReasoning: true,
+    toolCalling: true,
+  },
+  { id: "minimax-m2.7", name: "MiniMax M2.7", supportsReasoning: true, toolCalling: true },
+];
+
+export const cheaperinferenceProvider: RegistryEntry = {
+  id: "cheaperinference",
+  alias: "cinf",
+  format: "openai",
+  executor: "cheaperinference",
+  baseUrl: "https://api.cheaperinference.com/v1/chat/completions",
+  // The gateway serves a native, STATELESS /v1/responses endpoint alongside
+  // /v1/chat/completions. Consumed by CheaperInferenceExecutor.buildUrl for the
+  // models tagged targetFormat: "openai-responses" above.
+  responsesBaseUrl: "https://api.cheaperinference.com/v1/responses",
+  authType: "apikey",
+  authHeader: "bearer",
+  models: CHEAPERINFERENCE_MODELS,
+};

open-sse/executors/base.ts

@@ -34,6 +34,7 @@ import {
   resolveAccountKey,
   isFreeVariantModel,
 } from "../services/openrouterFreeWindow.ts";
+import { gateOutboundRequest } from "../services/wafRateLimit.ts";
 import type { PoolConfig } from "../services/sessionPool/types.ts";
 import type { Session } from "../services/sessionPool/session.ts";
 import { SessionPool } from "../services/sessionPool/sessionPool.ts";
@@ -45,6 +46,7 @@ import {
 } from "../services/apiKeyRotator.ts";
 import type { KeyHealth } from "../services/apiKeyRotator.ts";
 import { getOpenAICompatibleType, isClaudeCodeCompatible } from "../services/provider.ts";
+import { usesCcWireImage } from "../services/ccWireImageBuiltins.ts";
 import {
   runWithOnPersist,
   getRefreshLeadMs,
@@ -249,7 +251,10 @@ function collectThinkingConfigs(body: unknown): Array<Record<string, unknown>> {
   if (!body || typeof body !== "object") return [];
   const root = body as Record<string, unknown>;
   const configs: Array<Record<string, unknown>> = [];
-  const envelopes: unknown[] = [root.generationConfig, (root.request as Record<string, unknown> | undefined)?.generationConfig];
+  const envelopes: unknown[] = [
+    root.generationConfig,
+    (root.request as Record<string, unknown> | undefined)?.generationConfig,
+  ];
   for (const env of envelopes) {
     if (!env || typeof env !== "object") continue;
     const tc = (env as Record<string, unknown>).thinkingConfig;
@@ -445,6 +450,12 @@ export class BaseExecutor {
     );
   }
 
+  protected usesClaudeCodeProtocol(credentials: ProviderCredentials | null): boolean {
+    if (!isClaudeCodeCompatible(this.provider)) return false;
+    const format = this.resolveAlternate(credentials)?.format;
+    return format !== "openai" && format !== "openai-responses";
+  }
+
   /**
    * Resolve the effective API key via extra-keys round-robin rotation.
    * Mutates `credentials.providerSpecificData.selectedKeyId` on rotation.
@@ -589,6 +600,15 @@ export class BaseExecutor {
 
   // Intra-URL retry config: retry same URL before falling back to next node
   static readonly RETRY_CONFIG = { maxAttempts: 2, delayMs: 2000 };
+  // WAF (400 content-blocked) retry config: agentrouter.org's WAF is burst-sensitive
+  // and recovers after a short cooldown. Use exponential backoff with a higher
+  // starting delay than the generic 429 retry (which is 2s) because the WAF
+  // needs more time to clear its per-IP suspicion bucket.
+  static readonly WAF_RETRY_CONFIG = {
+    maxAttempts: 2,
+    delayMs: 1500,
+    backoffMultiplier: 2,
+  };
   // Timeout for receiving the initial upstream response headers. Once the response
   // starts streaming, STREAM_IDLE_TIMEOUT_MS / Undici bodyTimeout handle stalls.
   static FETCH_START_TIMEOUT_MS = FETCH_TIMEOUT_MS;
@@ -834,15 +854,16 @@ export class BaseExecutor {
         );
       }
 
-      const ccRequestDefaults = isClaudeCodeCompatible(this.provider)
+      const usesClaudeCodeProtocol = this.usesClaudeCodeProtocol(requestCredentials);
+      const fingerprintProvider =
+        usesCcWireImage(this.provider) && !usesClaudeCodeProtocol ? "codex" : this.provider;
+      const ccRequestDefaults = usesClaudeCodeProtocol
         ? getClaudeCodeCompatibleRequestDefaults(requestCredentials?.providerSpecificData)
         : {};
       const shouldForwardExtendedContext =
-        extendedContext &&
-        modelSupportsContext1mBeta(model) &&
-        !isClaudeCodeCompatible(this.provider);
+        extendedContext && modelSupportsContext1mBeta(model) && !usesClaudeCodeProtocol;
       const shouldForwardCcCompatibleContext1m =
-        isClaudeCodeCompatible(this.provider) &&
+        usesClaudeCodeProtocol &&
         ccRequestDefaults.context1m === true &&
         !modelHasNativeContext1m(model);
       if (shouldForwardExtendedContext || shouldForwardCcCompatibleContext1m) {
@@ -922,8 +943,8 @@ export class BaseExecutor {
           !activeCredentials?.apiKey;
 
         if (
-          this.provider === "claude" &&
-          (isClaudeCodeClient || hasClaudeOAuthToken) &&
+          ((this.provider === "claude" && (isClaudeCodeClient || hasClaudeOAuthToken)) ||
+            usesClaudeCodeProtocol) &&
           typeof transformedBody === "object" &&
           transformedBody !== null
         ) {
@@ -1200,6 +1221,11 @@ export class BaseExecutor {
             if (ccKeysLower.has(key.toLowerCase())) delete headers[key];
           }
           Object.assign(headers, ccHeaders);
+          if (usesCcWireImage(this.provider) && usesClaudeCodeProtocol) {
+            delete headers["Authorization"];
+            headers["x-api-key"] =
+              activeCredentials?.apiKey || activeCredentials?.accessToken || "";
+          }
           delete headers["X-Stainless-Helper-Method"];
 
           // OS/arch follow the host running the signed binary. Runtime version
@@ -1246,7 +1272,7 @@ export class BaseExecutor {
             // (tool_result must be in immediately next message).
             // Only apply for Claude/Claude-compatible — OpenAI allows results
             // spread across multiple subsequent messages.
-            const isClaude = this.provider === "claude" || isClaudeCodeCompatible(this.provider);
+            const isClaude = this.provider === "claude" || usesClaudeCodeProtocol;
             // For Claude, fixToolAdjacency may strip tool_use blocks whose
             // tool_result isn't in the next message; re-run fixToolPairs to
             // drop any tool_result orphaned by that strip (discussion #2410).
@@ -1265,7 +1291,7 @@ export class BaseExecutor {
         // at this final dispatch point — the single chokepoint every Claude
         // routing mode (grouped/raw/combo) and the native passthrough share,
         // before fingerprinting and CCH signing serialize the body.
-        if (this.provider === "claude" || isClaudeCodeCompatible(this.provider)) {
+        if (this.provider === "claude" || usesClaudeCodeProtocol) {
           enforceThinkingTemperature(transformedBody as Record<string, unknown>);
         }
 
@@ -1282,7 +1308,7 @@ export class BaseExecutor {
         // `contextEditingDisabled` (set by the 400-fallback) suppresses re-injection
         // when a fresh `transformedBody` is built for a retry/fallback URL.
         if (
-          (this.provider === "claude" || isClaudeCodeCompatible(this.provider)) &&
+          (this.provider === "claude" || usesClaudeCodeProtocol) &&
           contextEditing?.enabled &&
           !contextEditingDisabled
         ) {
@@ -1298,17 +1324,17 @@ export class BaseExecutor {
         let bodyString = JSON.stringify(transformedBody);
 
         const shouldFingerprint =
-          isCliCompatEnabled(this.provider) ||
+          isCliCompatEnabled(fingerprintProvider) ||
           (this.provider === "claude" && (isClaudeCodeClient || hasClaudeOAuthToken));
         if (shouldFingerprint) {
-          const fingerprinted = applyFingerprint(this.provider, headers, transformedBody);
+          const fingerprinted = applyFingerprint(fingerprintProvider, headers, transformedBody);
           finalHeaders = fingerprinted.headers;
           bodyString = fingerprinted.bodyString;
         }
 
         // CCH signing — replaces the cch=00000 placeholder in the billing
         // header with an xxHash64 integrity token over the serialized body.
-        if (isClaudeCodeCompatible(this.provider) || this.provider === "claude") {
+        if (usesClaudeCodeProtocol || this.provider === "claude") {
           bodyString = await signRequestBody(bodyString);
         }
 
@@ -1373,6 +1399,13 @@ export class BaseExecutor {
           recordFreeWindowAttempt(openrouterFreeWindowAccountKey);
         }
 
+        // WAF burst guard: agentrouter.org's content filter becomes more
+        // aggressive after rapid requests. Enforce a small inter-request gap
+        // to avoid tripping it. See open-sse/services/wafRateLimit.ts.
+        if (this.provider === "agentrouter") {
+          await gateOutboundRequest(`agentrouter:${url}`);
+        }
+
         let response = await fetchWithStartTimeout(url, fetchOptions);
 
         if (openrouterFreeWindowAccountKey) {
@@ -1396,7 +1429,7 @@ export class BaseExecutor {
             contextEditingDisabled = true;
             delete (transformedBody as Record<string, unknown>).context_management;
             let retryBody = JSON.stringify(transformedBody);
-            if (isClaudeCodeCompatible(this.provider) || this.provider === "claude") {
+            if (usesClaudeCodeProtocol || this.provider === "claude") {
               retryBody = await signRequestBody(retryBody);
             }
             log?.debug?.(
@@ -1435,7 +1468,7 @@ export class BaseExecutor {
             thinkingBudgetClampedMax = upstreamMax;
             if (clampNestedThinkingBudget(transformedBody, upstreamMax)) {
               let retryBody = JSON.stringify(transformedBody);
-              if (isClaudeCodeCompatible(this.provider) || this.provider === "claude") {
+              if (usesClaudeCodeProtocol || this.provider === "claude") {
                 retryBody = await signRequestBody(retryBody);
               }
               log?.info?.(
@@ -1466,7 +1499,7 @@ export class BaseExecutor {
             strippedFields.add(offending);
             delete (transformedBody as Record<string, unknown>)[offending];
             let retryBody = JSON.stringify(transformedBody);
-            if (isClaudeCodeCompatible(this.provider) || this.provider === "claude") {
+            if (usesClaudeCodeProtocol || this.provider === "claude") {
               retryBody = await signRequestBody(retryBody);
             }
             log?.debug?.(
@@ -1491,7 +1524,7 @@ export class BaseExecutor {
                   addParamToBlocklist(this.provider, autoLearned, model);
                   delete (transformedBody as Record<string, unknown>)[autoLearned];
                   let retryBody = JSON.stringify(transformedBody);
-                  if (isClaudeCodeCompatible(this.provider) || this.provider === "claude") {
+                  if (usesClaudeCodeProtocol || this.provider === "claude") {
                     retryBody = await signRequestBody(retryBody);
                   }
                   log?.info?.(
@@ -1510,6 +1543,34 @@ export class BaseExecutor {
           }
         }
 
+        // Intra-URL retry: agentrouter.org WAF returns 400 content-blocked
+        // intermittently (burst-sensitive, recovers after cooldown). Retry the
+        // same URL with exponential backoff before falling through to the
+        // 429/401/fallback chain. See docs/security/AGENTROUTER_WAF.md.
+        if (
+          !skipUpstreamRetry &&
+          response.status === HTTP_STATUS.BAD_REQUEST &&
+          (retryAttemptsByUrl[urlIndex] ?? 0) < BaseExecutor.WAF_RETRY_CONFIG.maxAttempts
+        ) {
+          const wafErrText = await response
+            .clone()
+            .text()
+            .catch(() => "");
+          if (/content[_-]blocked/i.test(wafErrText)) {
+            retryAttemptsByUrl[urlIndex] = (retryAttemptsByUrl[urlIndex] ?? 0) + 1;
+            const wafAttempt = retryAttemptsByUrl[urlIndex];
+            const wafBackoff = BaseExecutor.WAF_RETRY_CONFIG.delayMs *
+              Math.pow(BaseExecutor.WAF_RETRY_CONFIG.backoffMultiplier, wafAttempt - 1);
+            log?.debug?.(
+              "WAF_RETRY",
+              `400 content-blocked intra-retry ${wafAttempt}/${BaseExecutor.WAF_RETRY_CONFIG.maxAttempts} on ${url} — waiting ${wafBackoff}ms`
+            );
+            await new Promise((resolve) => setTimeout(resolve, wafBackoff));
+            urlIndex--; // re-run this urlIndex on the next loop iteration
+            continue;
+          }
+        }
+
         // Intra-URL retry: if 429 and we haven't exhausted per-URL retries, wait and retry the same URL
         if (
           !skipUpstreamRetry &&

open-sse/executors/cheaperinference.ts

@@ -0,0 +1,69 @@
+import { BaseExecutor, type ProviderCredentials } from "./base.ts";
+import { PROVIDERS } from "../config/constants.ts";
+import { getModelTargetFormat, PROVIDER_ID_TO_ALIAS } from "../config/providerModels.ts";
+
+/**
+ * CheaperInferenceExecutor — api.cheaperinference.com.
+ *
+ * The gateway is OpenAI-compatible on both surfaces, so everything else comes from
+ * BaseExecutor. Two provider-specific facts need handling (both measured against the
+ * live API on 2026-07-31, not inferred from docs):
+ *
+ *  1. `/v1/responses` is STATELESS and REQUIRES `store:false`. Omitting it returns
+ *     HTTP 400 ("This Responses-compatible endpoint is stateless. Send store=false…").
+ *     chatCore.ts deletes `store` for every provider except "openai" — a strip shared
+ *     by ~290 providers that must not be special-cased — so we re-add it here, after
+ *     that strip has run. A client-supplied `store:true` is overwritten rather than
+ *     forwarded: the endpoint cannot honour it, and forwarding would 400.
+ *
+ *  2. Chat and Responses live at DIFFERENT URLs (unlike providers that switch on a
+ *     path suffix). The per-model `targetFormat` registry tag is the single source of
+ *     truth for which surface a model uses — the same tag chatCore reads to translate
+ *     the body — so resolving the URL from it keeps URL and payload in lockstep.
+ *     Same pattern as executors/xai.ts (9router#2439).
+ */
+export class CheaperInferenceExecutor extends BaseExecutor {
+  constructor(provider = "cheaperinference") {
+    super(provider, PROVIDERS[provider]);
+  }
+
+  /**
+   * True when this model is served by the native /v1/responses endpoint.
+   *
+   * PROVIDER_MODELS is keyed by provider ALIAS ("cinf"), while PROVIDERS is keyed by
+   * provider ID ("cheaperinference") — so `this.provider` cannot be passed straight
+   * through the way executors/xai.ts does (there the alias equals the id, which hides
+   * the distinction). Resolve the alias first or every lookup silently returns null
+   * and every Responses request 400s upstream.
+   */
+  private usesResponsesEndpoint(model: string): boolean {
+    const alias = PROVIDER_ID_TO_ALIAS[this.provider] || this.provider;
+    return getModelTargetFormat(alias, model) === "openai-responses";
+  }
+
+  buildUrl(model: string, _stream: boolean, _urlIndex = 0): string {
+    if (this.usesResponsesEndpoint(model)) {
+      return this.config.responsesBaseUrl || this.config.baseUrl;
+    }
+    return this.config.baseUrl;
+  }
+
+  transformRequest(
+    model: string,
+    body: unknown,
+    stream: boolean,
+    credentials: ProviderCredentials
+  ): unknown {
+    const cleanedBody = super.transformRequest(model, body, stream, credentials);
+    if (!cleanedBody || typeof cleanedBody !== "object" || Array.isArray(cleanedBody)) {
+      return cleanedBody;
+    }
+    if (!this.usesResponsesEndpoint(model)) {
+      // Chat Completions rejects unknown params — never add `store` on that surface.
+      return cleanedBody;
+    }
+    return { ...(cleanedBody as Record<string, unknown>), store: false };
+  }
+}
+
+export default CheaperInferenceExecutor;

open-sse/executors/default.ts

@@ -55,6 +55,7 @@ import { forwardOpencodeClientHeaders } from "../utils/opencodeHeaders.ts";
 import { resolveZaiUrl } from "./default/zaiFormatOverride.ts";
 import { acquireNvidiaConcurrencySlot } from "./default/nvidiaConcurrencyGate.ts";
 import { resolveAlibabaProviderBaseUrl } from "@/shared/constants/alibabaProviderRegions";
+import { usesCcWireImage } from "../services/ccWireImageBuiltins.ts";
 
 import type { PoolConfig } from "../services/sessionPool/types.ts";
 
@@ -291,6 +292,10 @@ export class DefaultExecutor extends BaseExecutor {
       case "minimax":
       case "minimax-cn":
         return `${this.config.baseUrl}?beta=true`;
+      case "agentrouter":
+        return this.usesClaudeCodeProtocol(credentials)
+          ? `${this.config.baseUrl}?beta=true`
+          : this.config.baseUrl;
       case "gemini":
         return `${this.config.baseUrl}/${model}:${stream ? "streamGenerateContent?alt=sse" : "generateContent"}`;
       default: {
@@ -413,7 +418,7 @@ export class DefaultExecutor extends BaseExecutor {
         applyClineAuthHeaders(headers, credentials, effectiveKey, clientHeaders, false);
         break;
       default:
-        if (isClaudeCodeCompatible(this.provider)) {
+        if (this.usesClaudeCodeProtocol(credentials)) {
           const ccRequestDefaults = getClaudeCodeCompatibleRequestDefaults(
             credentials?.providerSpecificData
           );
@@ -423,6 +428,10 @@ export class DefaultExecutor extends BaseExecutor {
             credentials?.providerSpecificData?.ccSessionId,
             { redactThinking: ccRequestDefaults.redactThinking === true }
           );
+          if (usesCcWireImage(this.provider)) {
+            delete ccHeaders["Authorization"];
+            ccHeaders["x-api-key"] = effectiveKey || credentials.accessToken || "";
+          }
           // CC nodes are also anthropic-compatible-*, so honor operator custom
           // headers here (the early return skips the shared block below).
           applyCustomHeaders(ccHeaders, credentials.providerSpecificData?.customHeaders);

open-sse/executors/index.ts

@@ -50,6 +50,7 @@ import { PoeWebExecutor } from "./poe-web.ts";
 import { VeniceWebExecutor } from "./venice-web.ts";
 import { NotionWebExecutor } from "./notion-web.ts";
 import { V0VercelWebExecutor } from "./v0-vercel-web.ts";
+import { CheaperInferenceExecutor } from "./cheaperinference.ts";
 import { KimiWebExecutor } from "./kimi-web.ts";
 import { DoubaoWebExecutor } from "./doubao-web.ts";
 import { QwenWebExecutor } from "./qwen-web.ts";
@@ -165,6 +166,8 @@ const executors = {
   "kimi-coding": new KimiExecutor(), // Alias
   moonshot: new MoonshotExecutor(),
   kimi: new MoonshotExecutor("kimi"), // Hidden legacy Moonshot provider id
+  cheaperinference: new CheaperInferenceExecutor(),
+  cinf: new CheaperInferenceExecutor("cheaperinference"), // Alias
   "doubao-web": new DoubaoWebExecutor(),
   db: new DoubaoWebExecutor(), // Alias
   "qwen-web": new QwenWebExecutor(),
@@ -282,4 +285,5 @@ export { ZenmuxFreeExecutor } from "./zenmux-free.ts";
 export { HyperAgentExecutor } from "./hyperagent.ts";
 export { XaiExecutor } from "./xai.ts";
 export { MoonshotExecutor } from "./moonshot.ts";
+export { CheaperInferenceExecutor } from "./cheaperinference.ts";
 export { PromptQlExecutor } from "./promptql.ts";

open-sse/handlers/chatCore.ts

@@ -71,6 +71,7 @@ import { checkHeapPressureGuard } from "../utils/heapPressure.ts";
 import { normalizeHeaders } from "../utils/headers.ts";
 import { resolveChatCoreRequestFormat } from "./chatCore/requestFormat.ts";
 import { resolveChatCoreTargetFormat } from "./chatCore/targetFormat.ts";
+import { stripStore, usesClaudeBridge } from "./chatCore/agentRouterProtocol.ts";
 import { defaultClaudeToolType } from "./chatCore/claudeToolDefaults.ts";
 import { injectSystemPrompt, injectCustomSystemPrompt } from "../services/systemPrompt.ts";
 import { translateRequest, needsTranslation } from "../translator/index.ts";
@@ -341,7 +342,6 @@ import { OMNIROUTE_RESPONSE_HEADERS } from "@/shared/constants/headers";
 import { getClaudeCodeCompatibleRequestDefaults } from "@/lib/providers/requestDefaults";
 import {
   buildClaudeCodeCompatibleRequest,
-  isClaudeCodeCompatibleProvider,
   resolveClaudeCodeCompatibleSessionId,
 } from "../services/claudeCodeCompatible.ts";
 import { setGeminiThoughtSignatureMode } from "../services/geminiThoughtSignatureStore.ts";
@@ -751,6 +751,7 @@ export async function handleChatCore({
     provider,
     resolvedModel,
     apiFormat,
+    sourceFormat,
     customModelTargetFormat,
     providerSpecificData: credentials?.providerSpecificData,
   });
@@ -1920,7 +1921,7 @@ export async function handleChatCore({
 
   let translatedBody = body;
   const isClaudePassthrough = sourceFormat === FORMATS.CLAUDE && targetFormat === FORMATS.CLAUDE;
-  const isClaudeCodeCompatible = isClaudeCodeCompatibleProvider(provider);
+  const isClaudeCodeCompatible = usesClaudeBridge(provider, targetFormat, credentials);
   const isClaudeCodeSemanticPassthrough = isClaudeCodeSemanticPassthroughRequest({
     provider,
     sourceFormat,
@@ -2496,10 +2497,7 @@ export async function handleChatCore({
     log?.debug?.("PARAMS", `Renamed max_completion_tokens to max_tokens for ${model}`);
   }
 
-  // OpenAI's `store` parameter is not supported by most compatible providers and breaks them
-  if (provider !== "openai" && "store" in translatedBody) {
-    delete translatedBody.store;
-  }
+  stripStore(translatedBody, provider, targetFormat);
 
   // Chat clients may send stream_options.include_usage, but OpenAI Responses
   // upstreams (including Azure AI Foundry /responses) reject stream_options.

open-sse/handlers/chatCore/agentRouterProtocol.ts

@@ -0,0 +1,35 @@
+/**
+ * Per-request AgentRouter protocol decisions kept outside the chatCore orchestration god-file.
+ * AgentRouter exposes Claude Messages, OpenAI Chat, and OpenAI Responses as distinct upstream
+ * protocols, so its dynamic target format must override the connection's default Claude wire image.
+ */
+
+import { isClaudeCodeCompatibleProvider } from "../../services/claudeCodeCompatible.ts";
+import { FORMATS } from "../../translator/formats.ts";
+
+export function usesClaudeBridge(
+  provider: string,
+  targetFormat: string,
+  credentials: unknown
+): boolean {
+  const configuredTargetFormat = (
+    credentials as { providerSpecificData?: { targetFormat?: unknown } | null } | null | undefined
+  )?.providerSpecificData?.targetFormat;
+  const effectiveFormat = provider === "agentrouter" ? targetFormat : configuredTargetFormat;
+  return (
+    isClaudeCodeCompatibleProvider(provider) &&
+    effectiveFormat !== FORMATS.OPENAI &&
+    effectiveFormat !== FORMATS.OPENAI_RESPONSES
+  );
+}
+
+export function stripStore(
+  body: Record<string, unknown>,
+  provider: string,
+  targetFormat: string
+): void {
+  const supportsStore =
+    provider === "openai" ||
+    (provider === "agentrouter" && targetFormat === FORMATS.OPENAI_RESPONSES);
+  if (!supportsStore && "store" in body) delete body.store;
+}

open-sse/handlers/chatCore/executionCredentials.ts

@@ -5,8 +5,8 @@
  * Pure builder extracted from handleChatCore: derives the per-execution credentials object from the
  * resolved request context. Applies the native-Codex passthrough endpoint override, forces
  * apiType=responses (and the responses-upstream marker) for Azure AI Foundry / OCI when the model
- * routes to the OpenAI Responses format, and threads the Claude Code session id when present.
- * Side-effect-free; behaviour is byte-identical to the previous inline closure.
+ * routes to the OpenAI Responses format, synchronizes AgentRouter's per-request alternate protocol,
+ * and threads the Claude Code session id when present. Side-effect-free.
  */
 
 import { getKimiCodeStaticThinkingPolicy } from "../../config/providers/registry/kimi/coding/runtime.ts";
@@ -128,6 +128,16 @@ export function resolveExecutionCredentials(opts: {
     providerSpecificData.targetFormat = targetFormat;
   }
 
+  // AgentRouter exposes Claude, OpenAI Chat, and OpenAI Responses on distinct URLs with distinct
+  // auth schemes. Keep the executor's URL/header resolution synchronized with chatCore's resolved
+  // per-request protocol without persisting the inferred selection back to the connection.
+  if (
+    provider === "agentrouter" &&
+    (targetFormat === FORMATS.OPENAI || targetFormat === FORMATS.OPENAI_RESPONSES)
+  ) {
+    providerSpecificData.targetFormat = targetFormat;
+  }
+
   applyKimiExecutionMetadata(providerSpecificData, provider, targetFormat, modelInfo);
 
   const withApiType = {

open-sse/handlers/chatCore/targetFormat.ts

@@ -4,10 +4,11 @@
  *
  * Pure resolution of the provider alias + the upstream target format used to translate the request:
  * apiFormat==="responses" forces OpenAI Responses; otherwise the model's registry target format, then
- * the per-model custom override (#2905), then the provider default. Returns both `alias` (reused by
+ * the per-model custom override (#2905), then AgentRouter's matching inbound protocol when the
+ * connection has no explicit override, then the provider default. Returns both `alias` (reused by
  * the handler when stripping the `alias/` prefix off the upstream model id) and `targetFormat`.
- * Side-effect-free; byte-identical to the previous inline block. Sits alongside the other
- * request-setup resolvers (resolveChatCoreRequestSetup / resolveChatCoreRequestFormat).
+ * Side-effect-free; sits alongside the other request-setup resolvers
+ * (resolveChatCoreRequestSetup / resolveChatCoreRequestFormat).
  */
 
 import { PROVIDER_ID_TO_ALIAS, getModelTargetFormat } from "../../config/providerModels.ts";
@@ -18,16 +19,38 @@ export function resolveChatCoreTargetFormat(opts: {
   provider: string;
   resolvedModel: string;
   apiFormat: string | undefined;
+  sourceFormat?: string;
   customModelTargetFormat: string | undefined;
   providerSpecificData: unknown;
 }) {
-  const { provider, resolvedModel, apiFormat, customModelTargetFormat, providerSpecificData } = opts;
+  const {
+    provider,
+    resolvedModel,
+    apiFormat,
+    sourceFormat,
+    customModelTargetFormat,
+    providerSpecificData,
+  } = opts;
   const alias = PROVIDER_ID_TO_ALIAS[provider] || provider;
   const modelTargetFormat = getModelTargetFormat(alias, resolvedModel);
+  const explicitConnectionTargetFormat = (
+    providerSpecificData as { targetFormat?: unknown } | null | undefined
+  )?.targetFormat;
+  const inferredAgentRouterTargetFormat =
+    provider === "agentrouter" &&
+    !(typeof explicitConnectionTargetFormat === "string" && explicitConnectionTargetFormat) &&
+    (sourceFormat === FORMATS.OPENAI_RESPONSES ||
+      sourceFormat === FORMATS.OPENAI ||
+      sourceFormat === FORMATS.CLAUDE)
+      ? sourceFormat
+      : undefined;
   const targetFormat =
     apiFormat === "responses"
       ? FORMATS.OPENAI_RESPONSES
-      : modelTargetFormat || customModelTargetFormat || getTargetFormat(provider, providerSpecificData);
+      : modelTargetFormat ||
+        customModelTargetFormat ||
+        inferredAgentRouterTargetFormat ||
+        getTargetFormat(provider, providerSpecificData);
   return { alias, targetFormat };
 }
 

open-sse/services/combo/strategyDispatch.ts

@@ -0,0 +1,68 @@
+// open-sse/services/combo/strategyDispatch.ts
+// Runtime source of truth for the known-symbols combo gate (G1).
+//
+// HISTÓRICO: `scripts/check/check-known-symbols.ts` (seção 2) costumava descobrir quais
+// estratégias de roteamento têm branch de despacho lendo a fonte dos arquivos do combo e
+// extraindo literais `strategy === "..."` por regex. Isso quebra quando o despacho vira um
+// registry (R0.3): não há mais `strategy === "X"` para casar. Este módulo substitui essa
+// enumeração por regex-over-source por uma enumeração EXPLÍCITA em runtime, colada ao lado
+// do código de despacho real.
+//
+// Importamos as funções reais de ordenação/despacho (não apenas strings) para amarrar a
+// enumeração ao código vivo: se a maquinaria de despacho for reestruturada ou um módulo
+// quebrar, a importação falha no load do gate em vez de casar silenciosamente uma regex
+// obsoleta. As chaves em HANDLED_COMBO_STRATEGIES DEVEM casar exatamente o conjunto
+// canônico (ROUTING_STRATEGY_VALUES ∪ INTERNAL_ROUTING_STRATEGY_VALUES).
+//
+// Ao adicionar uma estratégia canônica, fie-a no despacho (aqui ou em combo.ts) e inclua-a
+// nesta lista; ao remover um branch, retire a entrada — o gate acusa qualquer divergência
+// nas duas direções (canonicalSemDespacho / despachoNaoCanonico).
+
+import { applyStrategyOrdering } from "./applyStrategyOrdering.ts";
+import { resolveAutoStrategyOrder } from "./resolveAutoStrategy.ts";
+import { tryFusionDispatch, tryPipelineDispatch } from "./dispatchPrelude.ts";
+import { resolveComboTargetPipeline } from "./targetResolution.ts";
+
+/**
+ * As funções reais que implementam o despacho/ordenação de estratégias. Referenciadas
+ * aqui para (a) provar ao gate que a maquinaria resolve e (b) servir de âncora viva para
+ * a enumeração abaixo — o registry que o R0.3 vai passar a consumir para o branching
+ * `strategy === ...` nasce destas mesmas funções.
+ */
+export const COMBO_STRATEGY_DISPATCH_LEAVES = {
+  applyStrategyOrdering,
+  resolveAutoStrategyOrder,
+  tryFusionDispatch,
+  tryPipelineDispatch,
+  resolveComboTargetPipeline,
+} as const;
+
+/**
+ * Conjunto exato de estratégias de roteamento que possuem implementação de despacho real.
+ *
+ * Cobertura esperada (em `main` do gate): este set ∪ IMPLICIT_DEFAULT_STRATEGIES deve
+ * igualar o canônico. Atualmente todas as 20 estratégias canônicas têm branch — então
+ * HANDLED_COMBO_STRATEGIES já contém as 20 e IMPLICIT_DEFAULT_STRATEGIES está vazio.
+ */
+export const HANDLED_COMBO_STRATEGIES: readonly string[] = [
+  "priority",
+  "weighted",
+  "round-robin",
+  "context-relay",
+  "fill-first",
+  "p2c",
+  "random",
+  "least-used",
+  "cost-optimized",
+  "reset-aware",
+  "reset-window",
+  "headroom",
+  "strict-random",
+  "auto",
+  "lkgp",
+  "context-optimized",
+  "cache-optimized",
+  "fusion",
+  "pipeline",
+  "quota-share",
+] as const;

open-sse/services/grokClearance.ts

@@ -1,3 +1,35 @@
+/**
+ * grokClearance.ts — gated browser-backed cf_clearance acquisition for
+ * grok-web (#8019).
+ *
+ * grok.com sits behind Cloudflare Enterprise, which pins `cf_clearance` to
+ * the client's IP+TLS+UA fingerprint. Pure cookie-replay from
+ * `grokTlsClient.ts` (TLS-impersonating fetch) cannot forge a fresh
+ * clearance from a datacenter egress that Cloudflare has already flagged —
+ * only a real browser solving the challenge natively can mint one bound to
+ * that egress's own fingerprint.
+ *
+ * This module reuses the EXISTING provider-agnostic browser pool
+ * (`browserPool.ts`, already live for claude-web + duckduckgo-web) rather
+ * than adding a new Turnstile solver — `claudeTurnstileSolver.ts` is
+ * claude.ai-specific and does not apply here.
+ *
+ * Opt-in only: gated behind `OMNIROUTE_BROWSER_POOL` / `WEB_COOKIE_USE_BROWSER`
+ * (the same env gate already used by claude-web.ts / duckduckgo-web.ts).
+ * With the gate off, `acquireFreshGrokClearance` is never called — the
+ * executor stays on the Step-1 `cloudflare_challenge` classification.
+ */
+
+import { acquireBrowserContext, type PooledContext } from "./browserPool.ts";
+
+const GROK_WARMUP_URL = "https://grok.com/";
+const GROK_COOKIE_DOMAIN = ".grok.com";
+const GROK_POOL_KEY = "grok-web";
+
+/**
+ * Reads the same opt-in gate as claude-web/duckduckgo-web
+ * (`WEB_COOKIE_USE_BROWSER` or `OMNIROUTE_BROWSER_POOL`). Off by default.
+ */
 export function shouldUseGrokBrowserBacked(): boolean {
   const flag = process.env.WEB_COOKIE_USE_BROWSER;
   if (flag === "1" || flag === "true" || flag === "on") return true;
@@ -5,16 +37,48 @@ export function shouldUseGrokBrowserBacked(): boolean {
   return poolFlag === "on" || poolFlag === "1" || poolFlag === "true";
 }
 
-let grokClearanceAcquireOverride: ((signal?: AbortSignal) => Promise<string | null>) | null = null;
+type AcquireGrokClearanceFn = (signal?: AbortSignal | null) => Promise<string | null>;
+
+// Test-only injection point — mirrors browserBackedChat.ts's
+// __setBrowserBackedChatOverrideForTesting pattern so unit tests can prove
+// the gating/wiring without launching a real browser (no chromium in CI).
+let acquireOverride: AcquireGrokClearanceFn | null = null;
 
 export function __setGrokClearanceAcquireOverrideForTesting(
-  fn: ((signal?: AbortSignal) => Promise<string | null>) | null,
+  fn: AcquireGrokClearanceFn | null
 ): void {
-  grokClearanceAcquireOverride = fn;
+  acquireOverride = fn;
 }
 
-export async function acquireFreshGrokClearance(signal?: AbortSignal): Promise<string | null> {
-  if (grokClearanceAcquireOverride) return grokClearanceAcquireOverride(signal);
-  const mod = await import("@omniroute/browser-pool");
-  return mod.acquireFreshGrokClearance(signal);
+async function readCfClearanceFromContext(pooled: PooledContext): Promise<string | null> {
+  const cookies = await pooled.context.cookies(GROK_WARMUP_URL);
+  const match = cookies.find((c) => c.name === "cf_clearance");
+  return match?.value || null;
+}
+
+async function acquireViaPool(): Promise<string | null> {
+  try {
+    const pooled = await acquireBrowserContext(GROK_POOL_KEY, {
+      cookieDomain: GROK_COOKIE_DOMAIN,
+      cookieString: null,
+      warmupUrl: GROK_WARMUP_URL,
+    });
+    return await readCfClearanceFromContext(pooled);
+  } catch {
+    return null;
+  }
+}
+
+/**
+ * Acquire a fresh `.grok.com` cf_clearance via the shared browser pool.
+ * Never throws — resolves to `null` on any failure so callers can fall
+ * through to the Cloudflare-challenge error rather than crash the request.
+ */
+export async function acquireFreshGrokClearance(signal?: AbortSignal | null): Promise<string | null> {
+  if (acquireOverride) return acquireOverride(signal);
+  try {
+    return await acquireViaPool();
+  } catch {
+    return null;
+  }
 }

open-sse/services/model.ts

@@ -120,7 +120,44 @@ for (const [aliasOrId, models] of Object.entries(PROVIDER_MODELS)) {
   }
 }
 const KNOWN_MODEL_IDS = new Set(MODEL_TO_PROVIDERS.keys());
-export const CODEX_NATIVE_UNPREFIXED_MODELS = new Set(["codex-auto-review"]);
+// Bare Codex CLI defaults must always route to the `codex` provider (chatgpt.com
+// OAuth) even when other providers that also catalog the model id (e.g.
+// `agentrouter`, `openai`) are active. The Codex cookie quota on the user's
+// account is the source of truth for capacity, and bare-id requests from
+// `codex` (CLI)/`Codex` (web) would otherwise silently fan out to whichever
+// provider won the inference race — leaving the user wondering why the
+// canonical ChatGPT subscription stopped working. Override per-request by
+// prefixing the model id (e.g. `agentrouter/gpt-5.6-sol`,
+// `openai/gpt-5.6-sol`) — the prefix path always wins.
+export const CODEX_NATIVE_UNPREFIXED_MODELS = new Set([
+  "codex-auto-review",
+  "gpt-5.6-sol",
+  "gpt-5.6-sol-ultra",
+  "gpt-5.6-sol-max",
+  "gpt-5.6-sol-xhigh",
+  "gpt-5.6-sol-high",
+  "gpt-5.6-sol-medium",
+  "gpt-5.6-sol-low",
+  "gpt-5.6-terra",
+  "gpt-5.6-terra-ultra",
+  "gpt-5.6-terra-max",
+  "gpt-5.6-terra-xhigh",
+  "gpt-5.6-terra-high",
+  "gpt-5.6-terra-medium",
+  "gpt-5.6-terra-low",
+  "gpt-5.6-luna",
+  "gpt-5.6-luna-max",
+  "gpt-5.6-luna-xhigh",
+  "gpt-5.6-luna-high",
+  "gpt-5.6-luna-medium",
+  "gpt-5.6-luna-low",
+  "gpt-5.5",
+  "gpt-5.5-xhigh",
+  "gpt-5.5-high",
+  "gpt-5.5-medium",
+  "gpt-5.5-low",
+  "gpt-5.3-codex-spark",
+]);
 
 interface ProviderConnectionLike {
   provider?: unknown;
@@ -534,7 +571,19 @@ async function resolveModelByProviderInference(modelId: string, extendedContext:
       getActiveSyncedProvidersForModel(modelId),
       getPreferClaudeCodeForUnprefixedClaudeModels(),
     ]);
-  const providers = getInferredProvidersForModel(modelId, activeSyncedProviders);
+  // #FIX: synced catalogs (populated from `/v1/models` per connection) can
+  // claim ownership of models the provider does not actually serve (e.g. a
+  // `kiro` upstream briefly advertising `claude-opus-5` before it was
+  // vendored into the registry). Without this filter the bare-routing path
+  // would forward traffic to providers that 404 on the upstream call.
+  // Auto-discovery still wins when no static registry entry exists for the
+  // model id — only entries that conflict with the static catalog are dropped.
+  const staticCatalogProviders = MODEL_TO_PROVIDERS.get(modelId) || [];
+  const validatedSyncedProviders =
+    staticCatalogProviders.length > 0
+      ? activeSyncedProviders.filter((p) => staticCatalogProviders.includes(p))
+      : activeSyncedProviders;
+  const providers = getInferredProvidersForModel(modelId, validatedSyncedProviders);
   const nonOpenAIProviders = providers.filter((p) => p !== "openai");
 
   // Bare model IDs from Codex CLI do not preserve OmniRoute's `cx/` prefix.

open-sse/services/wafRateLimit.ts

@@ -0,0 +1,76 @@
+/**
+ * wafRateLimit.ts — Burst guard for agentrouter.org upstream WAF.
+ *
+ * The agentrouter.org gateway runs a content-filter WAF that becomes more
+ * aggressive after bursts of requests from the same IP/key, returning
+ * `400 content-blocked` for requests that would normally pass. After ~5-10
+ * seconds of cooldown the filter relaxes again.
+ *
+ * To avoid tripping the WAF, we serialize outbound calls per provider and
+ * enforce a minimum inter-request gap. The defaults are conservative and
+ * meant to be a safety net — the upstream request rate from Claude Code is
+ * inherently low (one human-paced request at a time), so this guard should
+ * not affect normal traffic.
+ */
+
+import { log } from "../utils/logger.ts";
+
+interface BurstGuardState {
+  lastSentAt: number;
+}
+
+const state = new Map<string, BurstGuardState>();
+
+export interface WafRateLimitConfig {
+  minGapMs: number;
+}
+
+const DEFAULT_CONFIG: WafRateLimitConfig = {
+  // 500ms is enough to prevent the burst-sensitive WAF from activating
+  // while staying well below human perception of latency.
+  minGapMs: 500,
+};
+
+let config: WafRateLimitConfig = { ...DEFAULT_CONFIG };
+
+export function configureWafRateLimit(overrides: Partial<WafRateLimitConfig>): void {
+  config = { ...config, ...overrides };
+}
+
+export function getWafRateLimitConfig(): WafRateLimitConfig {
+  return { ...config };
+}
+
+/**
+ * Wait until at least `minGapMs` has passed since the last call to
+ * `gateOutboundRequest` for the same `bucketKey`. Safe to call from
+ * concurrent requests — the lock is held only for the sleep, not across
+ * the actual upstream fetch.
+ *
+ * @param bucketKey Stable identifier for the upstream (e.g. "agentrouter:url").
+ */
+export async function gateOutboundRequest(bucketKey: string): Promise<void> {
+  const now = Date.now();
+  const bucket = state.get(bucketKey);
+  if (!bucket) {
+    state.set(bucketKey, { lastSentAt: now });
+    return;
+  }
+  const elapsed = now - bucket.lastSentAt;
+  const wait = config.minGapMs - elapsed;
+  if (wait > 0) {
+    log?.debug?.(
+      "WAF_RATE_LIMIT",
+      `Throttling outbound to ${bucketKey} — waiting ${wait}ms (min gap ${config.minGapMs}ms)`
+    );
+    await new Promise((resolve) => setTimeout(resolve, wait));
+  }
+  state.set(bucketKey, { lastSentAt: Date.now() });
+}
+
+/**
+ * Reset all rate-limit state. Primarily for tests.
+ */
+export function resetWafRateLimit(): void {
+  state.clear();
+}

open-sse/utils/functionalGatewayMirrors.ts

@@ -0,0 +1,103 @@
+/**
+ * Functional gateway mirrors (`<gateway-alias>/<original-id>` mirror entries).
+ *
+ * /v1/models announces each model under its canonical owner provider
+ * (`deepseek/deepseek-v4-flash`). But the owner may have NO active credential
+ * while a passthrough gateway provider (e.g. agentrouter / openrouter) DOES and
+ * routes the same model. Discovery clients (omp, jcode, etc.) then see a model
+ * that fails on request, and never the route that works.
+ *
+ * This module synthesizes a mirror entry under the functional gateway alias:
+ *
+ *     <gateway-alias>/<original-id>     e.g. agentrouter/deepseek/deepseek-v4-flash
+ *
+ * The request path already resolves any known provider prefix
+ * (open-sse/services/model.ts::resolveProviderAlias), so the mirror is
+ * immediately routable with no request-side change. Pure synthesis over the
+ * already key-filtered list — no I/O.
+ */
+
+export const FUNCTIONAL_GATEWAY_MIRROR_SUFFIX = " (via ";
+
+export interface FunctionalGatewayMirrorsDeps {
+  /** Ordered list of passthrough gateway provider ids to consider as mirrors. */
+  gatewayProviderIds: string[];
+  /** True when `provider` is a passthrough gateway that can route arbitrary models. */
+  isGateway(provider: string): boolean;
+  /** Map a gateway provider id to its catalog alias (e.g. "command-code" -> "cmd"). */
+  gatewayAlias(provider: string): string;
+  /** True when `gatewayProvider` has an eligible connection that covers `modelId`. */
+  gatewayCovers(gatewayProvider: string, modelId: string): boolean;
+  /** True when `gatewayProvider` has an active credential/connection. */
+  gatewayHasConnection(gatewayProvider: string): boolean;
+  /** True when the canonical owner `provider` has an eligible connection for the model. */
+  canonicalOwnerHasConnection(provider: string): boolean;
+}
+
+interface GatewayMirrorCatalogEntry {
+  id?: unknown;
+  owned_by?: unknown;
+  root?: unknown;
+  name?: unknown;
+  display_name?: unknown;
+  [key: string]: unknown;
+}
+
+/**
+ * Append `<gatewayAlias>/<originalId>` mirror entries for every eligible model.
+ * Returns the original array reference unchanged when nothing is eligible.
+ */
+export function appendFunctionalGatewayMirrors<T extends GatewayMirrorCatalogEntry>(
+  models: T[],
+  deps: FunctionalGatewayMirrorsDeps
+): T[] {
+  if (!Array.isArray(models)) return models;
+
+  const aliases: T[] = [];
+  for (const model of models) {
+    const id = model.id;
+    if (typeof id !== "string" || id.length === 0) continue;
+
+    const slashIndex = id.indexOf("/");
+    if (slashIndex <= 0) continue; // no provider prefix to re-home
+    const owner = id.slice(0, slashIndex);
+    const modelId = id.slice(slashIndex + 1);
+    if (!modelId || modelId === id) continue;
+
+    // Skip if the canonical owner already has a working connection for this model.
+    if (deps.canonicalOwnerHasConnection(owner)) continue;
+
+    // Find a passthrough gateway that actually routes this model AND has a credential.
+    let chosenAlias: string | null = null;
+    let chosenProvider: string | null = null;
+    for (const gatewayProvider of deps.gatewayProviderIds) {
+      const alias = deps.gatewayAlias(gatewayProvider);
+      if (!alias || alias === owner) continue;
+      if (!deps.isGateway(gatewayProvider)) continue;
+      if (!deps.gatewayHasConnection(gatewayProvider)) continue;
+      if (!deps.gatewayCovers(gatewayProvider, modelId)) continue;
+      chosenAlias = alias;
+      chosenProvider = gatewayProvider;
+      break;
+    }
+    if (!chosenAlias || !chosenProvider) continue;
+
+    const aliasId = `${chosenAlias}/${id}`;
+    // Skip if the mirror already exists in the list.
+    if (models.some((m) => m.id === aliasId)) continue;
+    // Skip if the id already starts with this gateway alias (would double-prefix).
+    if (id.startsWith(`${chosenAlias}/`)) continue;
+
+    const label =
+      typeof model.name === "string" && model.name ? model.name : modelId;
+    aliases.push({
+      ...model,
+      id: aliasId,
+      root: id,
+      owned_by: chosenProvider,
+      display_name: `${label}${FUNCTIONAL_GATEWAY_MIRROR_SUFFIX}${chosenProvider})`,
+    } as T);
+  }
+
+  return aliases.length > 0 ? [...models, ...aliases] : models;
+}

open-sse/utils/passthroughTailProcessor.ts

@@ -3,6 +3,7 @@ import { parseSSEDataPayload } from "./streamHelpers.ts";
 import {
   backfillResponsesCompletedOutput,
   normalizeResponsesSseIds,
+  normalizeResponsesCompletedUsage,
   pushUniqueResponsesOutputItems,
   stringifyIdValue,
   stripResponsesLifecycleEcho,
@@ -174,13 +175,20 @@ function handleResponsesTailPayload(
   const outputPayload = textualToolCallBackfilled
     ? context.toResponsesCompletedWithToolCalls(parsed)
     : parsed;
+  const usageNormalized = normalizeResponsesCompletedUsage(outputPayload);
   const stripped = stripResponsesLifecycleEcho(outputPayload);
   const backfilled = backfillResponsesCompletedOutput(
     outputPayload,
     context.passthroughResponsesOutputItems
   );
 
-  if (stripped || backfilled || textualToolCallBackfilled || responsesIdsNormalized) {
+  if (
+    stripped ||
+    backfilled ||
+    textualToolCallBackfilled ||
+    responsesIdsNormalized ||
+    usageNormalized
+  ) {
     output = `data: ${JSON.stringify(outputPayload)}\n\n`;
   }
 

open-sse/utils/responsesStreamHelpers.ts

@@ -138,6 +138,49 @@ export function backfillResponsesCompletedOutput(
   return true;
 }
 
+/**
+ * Keep the terminal Responses payload compatible with strict clients such as Codex.
+ * Upstreams may expose only input/output counts (or omit usage entirely), while the
+ * client deserializer requires all three canonical token fields.
+ */
+export function normalizeResponsesCompletedUsage(parsed: unknown): boolean {
+  if (!parsed || typeof parsed !== "object" || Array.isArray(parsed)) return false;
+  const obj = parsed as JsonRecord;
+  if (obj.type !== "response.completed") return false;
+  if (!obj.response || typeof obj.response !== "object" || Array.isArray(obj.response)) {
+    return false;
+  }
+
+  const response = obj.response as JsonRecord;
+  const current =
+    response.usage && typeof response.usage === "object" && !Array.isArray(response.usage)
+      ? (response.usage as JsonRecord)
+      : {};
+  const finiteNumber = (value: unknown): number | null => {
+    const numeric = Number(value);
+    return Number.isFinite(numeric) ? numeric : null;
+  };
+  const inputTokens =
+    finiteNumber(current.input_tokens) ?? finiteNumber(current.prompt_tokens) ?? 0;
+  const outputTokens =
+    finiteNumber(current.output_tokens) ?? finiteNumber(current.completion_tokens) ?? 0;
+  const totalTokens = finiteNumber(current.total_tokens) ?? inputTokens + outputTokens;
+
+  const normalized: JsonRecord = {
+    ...current,
+    input_tokens: inputTokens,
+    output_tokens: outputTokens,
+  };
+  normalized.total_tokens = totalTokens;
+  const changed =
+    !response.usage ||
+    current.input_tokens !== inputTokens ||
+    current.output_tokens !== outputTokens ||
+    current.total_tokens !== totalTokens;
+  response.usage = normalized;
+  return changed;
+}
+
 const RESPONSES_LIFECYCLE_EVENT_TYPES = new Set([
   "response.created",
   "response.in_progress",

open-sse/utils/stream.ts

@@ -49,6 +49,7 @@ import {
 } from "../services/sessionManager.ts";
 import {
   backfillResponsesCompletedOutput,
+  normalizeResponsesCompletedUsage as normalizeUsage,
   normalizeResponsesSseIds,
   pushUniqueResponsesOutputItems,
   stringifyIdValue,
@@ -1581,11 +1582,13 @@ export function createSSEStream(options: StreamOptions = {}) {
                     parsed,
                     passthroughResponsesOutputItems
                   );
+                  const usageNormalized = normalizeUsage(parsed);
                   if (
                     stripped ||
                     backfilled ||
                     textualToolCallBackfilled ||
-                    responsesIdsNormalized
+                    responsesIdsNormalized ||
+                    usageNormalized
                   ) {
                     output = `data: ${JSON.stringify(parsed)}\n\n`;
                     injectedUsage = true;
@@ -2273,7 +2276,6 @@ export function createSSEStream(options: StreamOptions = {}) {
                 }
                 clientPayloadCollector.push(bufferedPayload);
 
-                // Normalize numeric IDs for final buffered data: chunk (same as transform path)
                 if (typeof bufferedPayload === "object" && !Array.isArray(bufferedPayload)) {
                   const flushedParsed = bufferedPayload as JsonRecord;
                   const flushedType =
@@ -2281,7 +2283,9 @@ export function createSSEStream(options: StreamOptions = {}) {
                   const isResponses = flushedType.startsWith("response.");
                   const isClaude = isClaudeEventPayload(flushedParsed);
                   if (isResponses) {
-                    if (normalizeResponsesSseIds(flushedParsed)) {
+                    const idsNormalized = normalizeResponsesSseIds(flushedParsed);
+                    const usageNormalized = normalizeUsage(flushedParsed);
+                    if (idsNormalized || usageNormalized) {
                       output = `data: ${JSON.stringify(flushedParsed)}\n\n`;
                     }
                   } else if (!isClaude) {

open-sse/utils/tlsClient.ts

@@ -40,7 +40,6 @@ interface FetchOptions {
   body?: unknown;
   redirect?: string;
   signal?: AbortSignal;
-  fingerprint?: { userAgent: string; secChUa: string; secChUaPlatform: string };
 }
 
 function normalizeHeaders(headers: HeadersInit | undefined): Record<string, string> | undefined {

package-lock.json

@@ -10,8 +10,7 @@
       "hasInstallScript": true,
       "license": "MIT",
       "workspaces": [
-        "open-sse",
-        "packages/browser-pool"
+        "open-sse"
       ],
       "dependencies": {
         "@aws-sdk/client-bedrock-runtime": "^3.1073.0",
@@ -3693,6 +3692,9 @@
       "cpu": [
         "arm"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3709,6 +3711,9 @@
       "cpu": [
         "arm64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3725,6 +3730,9 @@
       "cpu": [
         "ppc64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3741,6 +3749,9 @@
       "cpu": [
         "riscv64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3757,6 +3768,9 @@
       "cpu": [
         "s390x"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3773,6 +3787,9 @@
       "cpu": [
         "x64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3789,6 +3806,9 @@
       "cpu": [
         "arm64"
       ],
+      "libc": [
+        "musl"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3805,6 +3825,9 @@
       "cpu": [
         "x64"
       ],
+      "libc": [
+        "musl"
+      ],
       "license": "LGPL-3.0-or-later",
       "optional": true,
       "os": [
@@ -3821,6 +3844,9 @@
       "cpu": [
         "arm"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -3843,6 +3869,9 @@
       "cpu": [
         "arm64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -3865,6 +3894,9 @@
       "cpu": [
         "ppc64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -3887,6 +3919,9 @@
       "cpu": [
         "riscv64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -3909,6 +3944,9 @@
       "cpu": [
         "s390x"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -3931,6 +3969,9 @@
       "cpu": [
         "x64"
       ],
+      "libc": [
+        "glibc"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -3953,6 +3994,9 @@
       "cpu": [
         "arm64"
       ],
+      "libc": [
+        "musl"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -3975,6 +4019,9 @@
       "cpu": [
         "x64"
       ],
+      "libc": [
+        "musl"
+      ],
       "license": "Apache-2.0",
       "optional": true,
       "os": [
@@ -6283,10 +6330,6 @@
         "node": "^20.17.0 || >=22.9.0"
       }
     },
-    "node_modules/@omniroute/browser-pool": {
-      "resolved": "packages/browser-pool",
-      "link": true
-    },
     "node_modules/@omniroute/open-sse": {
       "resolved": "open-sse",
       "link": true
@@ -10696,6 +10739,9 @@
         "arm64"
       ],
       "dev": true,
+      "libc": [
+        "glibc"
+      ],
       "license": "MIT",
       "optional": true,
       "os": [
@@ -10713,6 +10759,9 @@
         "arm64"
       ],
       "dev": true,
+      "libc": [
+        "musl"
+      ],
       "license": "MIT",
       "optional": true,
       "os": [
@@ -10730,6 +10779,9 @@
         "x64"
       ],
       "dev": true,
+      "libc": [
+        "glibc"
+      ],
       "license": "MIT",
       "optional": true,
       "os": [
@@ -10747,6 +10799,9 @@
         "x64"
       ],
       "dev": true,
+      "libc": [
+        "musl"
+      ],
       "license": "MIT",
       "optional": true,
       "os": [
@@ -12839,6 +12894,9 @@
         "arm"
       ],
       "dev": true,
+      "libc": [
+        "glibc"
+      ],
       "optional": true,
       "os": [
         "linux"
@@ -12852,6 +12910,9 @@
         "arm"
       ],
       "dev": true,
+      "libc": [
+        "musl"
+      ],
       "optional": true,
       "os": [
         "linux"
@@ -12865,6 +12926,9 @@
         "arm64"
       ],
       "dev": true,
+      "libc": [
+        "glibc"
+      ],
       "optional": true,
       "os": [
         "linux"
@@ -12878,6 +12942,9 @@
         "arm64"
       ],
       "dev": true,
+      "libc": [
+        "musl"
+      ],
       "optional": true,
       "os": [
         "linux"
@@ -12891,6 +12958,9 @@
         "x64"
       ],
       "dev": true,
+      "libc": [
+        "glibc"
+      ],
       "optional": true,
       "os": [
         "linux"
@@ -12904,6 +12974,9 @@
         "x64"
       ],
       "dev": true,
+      "libc": [
+        "musl"
+      ],
       "optional": true,
       "os": [
         "linux"
@@ -37012,33 +37085,6 @@
         "safe-regex": "^2.1.1",
         "smol-toml": "1.7.1"
       }
-    },
-    "packages/browser-pool": {
-      "name": "@omniroute/browser-pool",
-      "version": "0.1.0",
-      "dependencies": {
-        "playwright": "1.61.1"
-      },
-      "devDependencies": {
-        "@types/node": "^22"
-      }
-    },
-    "packages/browser-pool/node_modules/@types/node": {
-      "version": "22.20.1",
-      "resolved": "https://registry.npmjs.org/@types/node/-/node-22.20.1.tgz",
-      "integrity": "sha512-EANqOCF9QFyra+4pfxUcX9STKJpCLjMbObVzljIJomAWSnuSIEAvyzEU53GaajbXJEgdh0iEcPL+DGvpUd4k1Q==",
-      "dev": true,
-      "license": "MIT",
-      "dependencies": {
-        "undici-types": "~6.21.0"
-      }
-    },
-    "packages/browser-pool/node_modules/undici-types": {
-      "version": "6.21.0",
-      "resolved": "https://registry.npmjs.org/undici-types/-/undici-types-6.21.0.tgz",
-      "integrity": "sha512-iwDZqg0QAGrg9Rav5H4n0M64c3mkR59cJ6wQp+7C4nI0gsmExaedaYLNO44eT4AtBBwjbTiGPMlt2Md0T9H9JQ==",
-      "dev": true,
-      "license": "MIT"
     }
   }
 }

package.json

@@ -1,7 +1,7 @@
 {
   "name": "omniroute",
   "version": "3.8.50",
-  "description": "Unified AI router with 160+ providers, RTK+Caveman compression, auto fallback, MCP/A2A, desktop, PWA, and OpenAI-compatible APIs.",
+  "description": "Unified AI router with 290 providers, RTK+Caveman compression, auto fallback, MCP/A2A, desktop, PWA, and OpenAI-compatible APIs.",
   "type": "module",
   "bin": {
     "omniroute": "bin/omniroute.mjs",
@@ -47,8 +47,7 @@
     "!**/*.spec.tsx"
   ],
   "workspaces": [
-    "open-sse",
-    "packages/browser-pool"
+    "open-sse"
   ],
   "engines": {
     "node": ">=22.22.2 <23 || >=24.0.0 <27"
@@ -111,6 +110,8 @@
     "test:unit:ci": "cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=4 tests/unit/*.test.ts \"tests/unit/{api,auth,authz,build,cli,cli-helper,combo,compression,correctness,cors,db,db-adapters,docs,gamification,guardrails,lib,mcp,memory,runtime,security,services,settings,shared,ui,usage}/**/*.test.ts\" \"tests/unit/**/*.test.mjs\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=4 \"tests/unit/dashboard/**/*.test.ts\" && npm run test:unit:serial",
     "test:unit:ci:shard": "cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=4096 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=4 --test-shard=$TEST_SHARD tests/unit/*.test.ts \"tests/unit/{api,auth,authz,build,cli,cli-helper,combo,compression,correctness,cors,db,db-adapters,docs,gamification,guardrails,lib,mcp,memory,runtime,security,services,settings,shared,ui,usage}/**/*.test.ts\" \"tests/unit/**/*.test.mjs\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=4096 --import tsx --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=4 --test-shard=$TEST_SHARD \"tests/unit/dashboard/**/*.test.ts\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=4096 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=1 --test-shard=$TEST_SHARD \"tests/unit/serial/**/*.test.ts\"",
     "test:unit:fast": "cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-isolation=none tests/unit/*.test.ts \"tests/unit/{api,auth,authz,build,cli,cli-helper,combo,compression,correctness,cors,db,db-adapters,docs,gamification,guardrails,lib,mcp,memory,runtime,security,services,settings,shared,ui,usage}/**/*.test.ts\" \"tests/unit/**/*.test.mjs\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-isolation=none \"tests/unit/dashboard/**/*.test.ts\" && npm run test:unit:serial",
+    "test:scoped": "bash scripts/quality/test-scoped.sh",
+    "test:scoped:staged": "bash scripts/quality/test-scoped.sh --staged",
     "test:unit:shard": "concurrently --kill-others-on-fail -n s1,s2 \"npm:test:unit:shard:1\" \"npm:test:unit:shard:2\"",
     "test:unit:shard:1": "cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=10 --test-shard=1/2 tests/unit/*.test.ts \"tests/unit/{api,auth,authz,build,cli,cli-helper,combo,compression,correctness,cors,db,db-adapters,docs,gamification,guardrails,lib,mcp,memory,runtime,security,services,settings,shared,ui,usage}/**/*.test.ts\" \"tests/unit/**/*.test.mjs\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=10 --test-shard=1/2 \"tests/unit/dashboard/**/*.test.ts\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=1 --test-shard=1/2 \"tests/unit/serial/**/*.test.ts\"",
     "test:unit:shard:2": "cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=10 --test-shard=2/2 tests/unit/*.test.ts \"tests/unit/{api,auth,authz,build,cli,cli-helper,combo,compression,correctness,cors,db,db-adapters,docs,gamification,guardrails,lib,mcp,memory,runtime,security,services,settings,shared,ui,usage}/**/*.test.ts\" \"tests/unit/**/*.test.mjs\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=10 --test-shard=2/2 \"tests/unit/dashboard/**/*.test.ts\" && cross-env DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx/esm --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit --test-concurrency=1 --test-shard=2/2 \"tests/unit/serial/**/*.test.ts\"",
@@ -304,6 +305,7 @@
     "smol-toml": "1.7.1",
     "socks": "^2.8.7",
     "sql.js": "^1.14.1",
+    "sqlite-vec": "^0.1.9",
     "tailwind-merge": "^3.6.0",
     "tsx": "^4.23.0",
     "undici": "^8.3.0",
@@ -323,8 +325,7 @@
     "js-tiktoken": "^1.0.20",
     "keytar": "^7.9.0",
     "tls-client-node": "^0.2.0",
-    "wreq-js": "^2.3.1",
-    "sqlite-vec": "^0.1.9"
+    "wreq-js": "^2.3.1"
   },
   "devDependencies": {
     "@axe-core/playwright": "^4.11.3",

packages/browser-pool/package.json

@@ -1,15 +0,0 @@
-{
-  "name": "@omniroute/browser-pool",
-  "version": "0.1.0",
-  "private": true,
-  "description": "Optional browser pool service for OmniRoute — CloakBrowser and Playwright-backed chat",
-  "type": "module",
-  "main": "./src/index.ts",
-  "types": "./src/index.ts",
-  "dependencies": {
-    "playwright": "1.61.1"
-  },
-  "devDependencies": {
-    "@types/node": "^22"
-  }
-}

packages/browser-pool/src/index.ts

@@ -1,37 +0,0 @@
-/**
- * @omniroute/browser-pool — Optional browser pool for Playwright-backed
- * executor support (claude-web, duckduckgo-web, grok).
- *
- * Core stubs dynamically import this package at runtime. When the package
- * is not installed, the stubs degrade gracefully (fallback or error).
- */
-
-// ── Re-exports from browserPool ──────────────────────────────────────────
-export {
-  acquireBrowserContext,
-  releaseBrowserContext,
-  getBrowserPoolMetrics,
-  readPageResponseBody,
-  openPage,
-  shutdownPool,
-  setProxyResolver,
-  __resetBrowserPoolMetricsForTest,
-} from "./services/browserPool.ts";
-
-export type { BrowserPoolContextOptions, BrowserPoolMetrics, PooledContext } from "./interfaces.ts";
-
-// ── Re-exports from browserBackedChat ────────────────────────────────────
-export {
-  browserBackedChat,
-  startBrowserWarmup,
-  getFreshCookiesWithWarmup,
-} from "./services/browserBackedChat.ts";
-
-// ── Re-exports from grokClearance ─────────────────────────────────────────
-export {
-  getCachedCookies,
-  setCachedCookies,
-  clearCookieCache,
-} from "./services/browserBackedChat.ts";
-
-export { shouldUseGrokBrowserBacked, acquireFreshGrokClearance } from "./services/grokClearance.ts";

packages/browser-pool/src/interfaces.ts

@@ -1,94 +0,0 @@
-/**
- * interfaces.ts — Shared type definitions for @omniroute/browser-pool.
- *
- * These types are used by both the package entry and the core stubs.
- * The core stubs re-export them so existing import paths remain stable.
- */
-
-import type { BrowserContext, Page } from "playwright";
-
-// ── Browser pool ───────────────────────────────────────
-
-export interface BrowserPoolContextOptions {
-  cookieDomain: string;
-  cookieString?: string | null;
-  warmupUrl?: string | null;
-  userAgent?: string;
-  locale?: string;
-  timezone?: string;
-  preferCloakbrowser?: boolean;
-  /** Time (ms) to wait for the warmup page to be ready. */
-  waitFor?: number;
-}
-
-export interface PooledContext {
-  id: string;
-  context: BrowserContext;
-  warmupPage: Page | null;
-  lastUsed: number;
-  isStealth: boolean;
-}
-
-export interface BrowserPoolMetrics {
-  browserLaunches: number;
-  browserLaunchFailures: number;
-  contextsCreated: number;
-  contextsReused: number;
-  contextsEvicted: number;
-  contextsReleased: number;
-  contextCreateFailures: number;
-  shutdowns: number;
-  lastShutdownReason: string | null;
-}
-
-// ── Browser-backed chat ────────────────────────────────
-
-export interface BrowserBackedChatRequest {
-  /** Pool key — typically a provider id like "duckduckgo-web" or
-   *  "claude-web", optionally suffixed by user/account id. */
-  poolKey: string;
-  /** Chat URL the page should submit to (captured via waitForResponse). */
-  chatUrl: string;
-  /** Chat page URL to navigate to before typing. */
-  chatPageUrl: string;
-  /** The text the user wants to send. */
-  userMessage: string;
-  /** Cookie string (raw) to inject into the browser context. */
-  cookieString?: string | null;
-  /** Cookie domain (used together with cookieString). */
-  cookieDomain?: string;
-  /** Domain for the page's fetch to identify the chat endpoint. */
-  chatUrlMatchDomain: string;
-  /** User-Agent string for the browser context. */
-  userAgent?: string;
-  /** Locale (BCP 47). Defaults to en-US. */
-  locale?: string;
-  /** IANA timezone. Defaults to America/New_York. */
-  timezone?: string;
-  /** Selector for the chat input. */
-  inputSelector: string;
-  /** Selector for the submit button (optional — falls back to Enter). */
-  submitButtonSelector?: string;
-  /** Wait after submit for SSE/JSON to arrive. Default 15 seconds. */
-  postSubmitWaitMs?: number;
-  /** Optional AbortSignal. Cancels navigation/submit. */
-  signal?: AbortSignal | null;
-  /** Reuse the same context across requests. Default true. */
-  reuseContext?: boolean;
-}
-
-export interface BrowserBackedChatTiming {
-  acquireContextMs: number;
-  navigateMs: number;
-  submitMs: number;
-  captureResponseMs: number;
-  totalMs: number;
-}
-
-export interface BrowserBackedChatResult {
-  status: number;
-  contentType: string | null;
-  body: Buffer;
-  isStealth: boolean;
-  timing: BrowserBackedChatTiming;
-}

packages/browser-pool/src/services/browserBackedChat.ts

@@ -1,461 +0,0 @@
-/**
- * browserBackedChat.ts — Full browser-backed chat interaction for @omniroute/browser-pool.
- *
- * Opens a page on a shared browser context, navigates to the provider's
- * chat page, types the user's message, clicks Send, and returns the
- * upstream SSE/JSON response body as a structured result.
- *
- * Providers using this path: duckduckgo-web, claude-web.
- *
- * The browser solves the provider's challenge natively (VQD, Cloudflare
- * Turnstile, etc.) by computing real DOM measurement values. The
- * Node-side challenge solver still runs as a first-line best-effort;
- * this module is the fallback.
- */
-
-import { Buffer } from "node:buffer";
-import {
-  acquireBrowserContext,
-  openPage,
-  readPageResponseBody,
-  releaseBrowserContext,
-} from "./browserPool.ts";
-import type {
-  PooledContext,
-  BrowserBackedChatRequest,
-  BrowserBackedChatResult,
-} from "../interfaces.ts";
-
-// Safety constants
-const MAX_RESPONSE_BYTES = 10 * 1024 * 1024; // 10 MB
-
-// Cookie cache constants
-const COOKIE_CACHE_TTL_MS = 5 * 60 * 1000; // Cache fresh cookies for 5 minutes
-const COOKIE_POLL_INTERVAL_MS = 500; // Poll for cookies every 500ms
-const COOKIE_POLL_TIMEOUT_MS = 5000; // Max poll time for cookies
-
-// Cookie cache — avoids repeated browser launches when cookies are still valid
-interface CachedCookies {
-  cookieString: string;
-  expiresAt: number;
-  domain: string;
-}
-const cookieCache = new Map<string, CachedCookies>();
-
-export function getCachedCookies(domain: string): string | null {
-  const cached = cookieCache.get(domain);
-  if (cached && Date.now() < cached.expiresAt) return cached.cookieString;
-  cookieCache.delete(domain);
-  return null;
-}
-
-export function setCachedCookies(domain: string, cookieString: string, ttlMs?: number): void {
-  cookieCache.set(domain, {
-    cookieString,
-    expiresAt: Date.now() + (ttlMs ?? COOKIE_CACHE_TTL_MS),
-    domain,
-  });
-}
-
-export function clearCookieCache(): void {
-  cookieCache.clear();
-}
-
-// Dedup pending cookie refreshes per pool key
-const pendingRefreshes = new Map<string, Promise<string | null>>();
-
-/** Sanitize an error message for safe JSON transport. */
-const MAX_ERROR_LEN = 512;
-function sanitizeErrorMessage(message: unknown): string {
-  let str = typeof message === "string" ? message : String(message ?? "");
-  if (str.length > MAX_ERROR_LEN) str = str.slice(0, MAX_ERROR_LEN);
-  const nl = str.indexOf("\n");
-  if (nl >= 0) str = str.slice(0, nl);
-  return str.replace(/[^ -~]/g, "").trim();
-}
-
-/** Wait N milliseconds, abortable via signal. */
-async function waitWithSignal(ms: number, signal?: AbortSignal | null): Promise<void> {
-  if (signal?.aborted) throw new DOMException("Aborted", "AbortError");
-  return new Promise<void>((resolve, reject) => {
-    const onAbort = () => {
-      clearTimeout(timer);
-      reject(new DOMException("Aborted", "AbortError"));
-    };
-    const timer = setTimeout(() => {
-      signal?.removeEventListener("abort", onAbort);
-      resolve();
-    }, ms);
-    signal?.addEventListener("abort", onAbort, { once: true });
-  });
-}
-
-/**
- * waitForCookiesWithPolling — Poll for cookies every 500ms up to 5s.
- * Returns as soon as challenge cookies appear, instead of always
- * waiting the full timeout. Saves 1-4s when anti-bot resolves quickly.
- */
-async function waitForCookiesWithPolling(
-  context: import("playwright").BrowserContext,
-  cookieDomain: string,
-  signal: AbortSignal | null
-): Promise<string | null> {
-  const deadline = Date.now() + COOKIE_POLL_TIMEOUT_MS;
-  while (Date.now() < deadline) {
-    if (signal?.aborted) throw new DOMException("Aborted", "AbortError");
-    const cookies = await context.cookies(cookieDomain);
-    const cookieString = cookies.map((c) => `${c.name}=${c.value}`).join("; ");
-    if (cookieString) return cookieString;
-    const remaining = deadline - Date.now();
-    if (remaining <= 0) break;
-    await waitWithSignal(Math.min(COOKIE_POLL_INTERVAL_MS, remaining), signal);
-  }
-  return null;
-}
-
-/**
- * doCookieRefreshOnContext — Run cookie extraction on an already-acquired
- * browser context. Opens a temporary page, navigates to the chat URL,
- * polls for cookies, and returns the result.
- * NOTE: Does NOT pass AbortSignal to Playwright methods — signals are
- * handled via waitWithSignal wrapping instead.
- */
-async function doCookieRefreshOnContext(
-  pooled: PooledContext,
-  chatPageUrl: string,
-  cookieDomain: string,
-  signal: AbortSignal | null
-): Promise<string | null> {
-  const page = await openPage(pooled);
-  try {
-    await page.goto(chatPageUrl, {
-      waitUntil: "domcontentloaded",
-      timeout: 60000,
-    });
-    return await waitForCookiesWithPolling(pooled.context, cookieDomain, signal);
-  } catch (err) {
-    if (err instanceof DOMException && err.name === "AbortError") throw err;
-    return null;
-  } finally {
-    await page.close().catch(() => {});
-  }
-}
-
-/**
- * Match a URL against a chat URL template, allowing a single dynamic
- * id segment (PLACEHOLDER) in the template.
- */
-function chatUrlMatcher(u: string, matchDomain: string, chatUrl: string): boolean {
-  if (u === chatUrl) return true;
-  let parsed: URL;
-  let chatParsed: URL;
-  try {
-    parsed = new URL(u);
-    chatParsed = new URL(chatUrl);
-  } catch {
-    return false;
-  }
-  if (!parsed.host.endsWith(matchDomain)) return false;
-  const chatSeg = chatParsed.pathname.split("/").filter(Boolean);
-  const reqSeg = parsed.pathname.split("/").filter(Boolean);
-  if (chatSeg.length < 2 || reqSeg.length !== chatSeg.length) return false;
-  let allowedDynamic = 1;
-  for (let i = 0; i < chatSeg.length; i++) {
-    if (chatSeg[i] === reqSeg[i]) continue;
-    if (chatSeg[i] === "PLACEHOLDER" && allowedDynamic > 0) {
-      allowedDynamic--;
-      continue;
-    }
-    return false;
-  }
-  return true;
-}
-
-/** Resolve a unique pool key; when reuseContext is false, create a unique key. */
-async function settlePoolKey(
-  requestedKey: string,
-  reuseContext: boolean
-): Promise<{ key: string; acquired: boolean }> {
-  if (reuseContext) return { key: requestedKey, acquired: true };
-  return {
-    key: `${requestedKey}-${Date.now()}-${Math.random().toString(36).slice(2, 8)}`,
-    acquired: false,
-  };
-}
-
-// ── Cookie refresh helpers ──────────────────────────
-
-/**
- * doRefresh — Acquire a fresh browser context, navigate to the
- * chat page, and poll for cookies. Returns the cookie string
- * or null on failure.
- * NOTE: Does NOT pass AbortSignal to Playwright methods — signals
- * are handled via waitWithSignal wrapping instead.
- */
-async function doRefresh(options: {
-  chatPageUrl: string;
-  cookieDomain: string;
-  poolKey: string;
-  signal: AbortSignal | null;
-}): Promise<string | null> {
-  const pooled = await acquireBrowserContext(options.poolKey + "-refresh", {
-    cookieDomain: options.cookieDomain,
-    cookieString: null,
-    warmupUrl: options.chatPageUrl,
-  });
-  const page = await openPage(pooled);
-  try {
-    await page.goto(options.chatPageUrl, {
-      waitUntil: "domcontentloaded",
-      timeout: 60000,
-    });
-    return await waitForCookiesWithPolling(pooled.context, options.cookieDomain, options.signal);
-  } catch (err) {
-    if (err instanceof DOMException && err.name === "AbortError") throw err;
-    return null;
-  } finally {
-    await page.close().catch(() => {});
-    // release context — we got the cookies
-    setTimeout(() => {
-      releaseBrowserContext(options.poolKey + "-refresh").catch(() => {});
-    }, 1000);
-  }
-}
-
-/**
- * refreshCookiesViaBrowser — Refresh cookies using a browser context.
- * Uses pendingRefreshes dedup so concurrent requests share one browser launch.
- * NOTE: Override check (httpOverride) is handled in the core stub — this
- * package version always attempts browser cookie refresh.
- */
-async function refreshCookiesViaBrowser(
-  chatUrl: string,
-  chatPageUrl: string,
-  cookieDomain: string,
-  poolKey: string,
-  signal: AbortSignal | null
-): Promise<string | null> {
-  const pending = pendingRefreshes.get(poolKey);
-  if (pending) return pending;
-  const promise = doRefresh({ chatPageUrl, cookieDomain, poolKey, signal });
-  pendingRefreshes.set(poolKey, promise);
-  try {
-    return await promise;
-  } finally {
-    pendingRefreshes.delete(poolKey);
-  }
-}
-
-/**
- * startBrowserWarmup — Pre-warm a browser context for the given pool key.
- * This is a fire-and-forget operation: errors are caught and ignored.
- * The warmup page serves as a readiness indicator — we open a page in the
- * pooled context to force early navigation before the actual request.
- */
-export async function startBrowserWarmup(
-  poolKey: string,
-  chatPageUrl: string,
-  cookieDomain: string,
-  signal: AbortSignal | null
-): Promise<void> {
-  if (process.env.OMNIROUTE_BROWSER_POOL === "off") return;
-  const pooled = await acquireBrowserContext(poolKey, {
-    cookieDomain,
-    cookieString: null,
-    warmupUrl: chatPageUrl,
-    waitFor: 2000,
-  });
-  // Warmup: open a page in the pooled context — this can happen in parallel
-  openPage(pooled).catch(() => {});
-}
-
-/**
- * getFreshCookiesWithWarmup — Try cached cookies first; if none, start
- * a browser warmup in parallel with a cookie refresh. Returns cookie string
- * or null. Caches successful results.
- */
-export async function getFreshCookiesWithWarmup(
-  chatUrl: string,
-  chatPageUrl: string,
-  cookieDomain: string,
-  poolKey: string,
-  signal: AbortSignal | null
-): Promise<string | null> {
-  // Try cached cookies first
-  const cached = getCachedCookies(cookieDomain);
-  if (cached) return cached;
-
-  // Start warmup in parallel with refresh
-  const warmup = startBrowserWarmup(poolKey, chatPageUrl, cookieDomain, signal);
-  const fresh = await refreshCookiesViaBrowser(chatUrl, chatPageUrl, cookieDomain, poolKey, signal);
-  // Await warmup (errors are non-fatal)
-  await warmup.catch(() => {});
-  if (fresh) {
-    setCachedCookies(cookieDomain, fresh);
-    return fresh;
-  }
-  return null;
-}
-
-// ── Main entry point ───────────────────────────────────
-
-export async function browserBackedChat(
-  req: BrowserBackedChatRequest
-): Promise<BrowserBackedChatResult> {
-  const t0 = Date.now();
-  const {
-    poolKey,
-    chatUrl,
-    chatPageUrl,
-    userMessage,
-    cookieString,
-    cookieDomain,
-    chatUrlMatchDomain,
-    userAgent,
-    locale,
-    timezone,
-    inputSelector,
-    submitButtonSelector,
-    postSubmitWaitMs = 15000,
-    signal,
-    reuseContext = true,
-  } = req;
-
-  const { key, acquired: reuseAcquired } = await settlePoolKey(poolKey, reuseContext);
-  const tAcquireStart = Date.now();
-  const pooled: PooledContext = await acquireBrowserContext(key, {
-    cookieDomain: cookieDomain || chatUrlMatchDomain,
-    cookieString: cookieString || null,
-    warmupUrl: chatPageUrl,
-    userAgent,
-    locale,
-    timezone,
-  });
-  const acquireContextMs = Date.now() - tAcquireStart;
-
-  const page = await openPage(pooled);
-  try {
-    const tNavStart = Date.now();
-    await page.goto(chatPageUrl, {
-      waitUntil: "domcontentloaded",
-      timeout: 60000,
-    });
-    const navigateMs = Date.now() - tNavStart;
-
-    const inputLocator = page.locator(inputSelector).first();
-    await inputLocator.waitFor({ state: "visible", timeout: 10000 });
-    await waitWithSignal(800, signal);
-
-    const responsePromise = page.waitForResponse(
-      (r) =>
-        r.request().method() === "POST" && chatUrlMatcher(r.url(), chatUrlMatchDomain, chatUrl),
-      { timeout: 30000 }
-    );
-
-    let abortListener: (() => void) | undefined;
-    const signalPromise = signal
-      ? new Promise<never>((_, reject) => {
-          if (signal.aborted) return reject(new DOMException("Aborted", "AbortError"));
-          abortListener = () => reject(new DOMException("Aborted", "AbortError"));
-          signal.addEventListener("abort", abortListener, { once: true });
-        })
-      : null;
-
-    if (submitButtonSelector) {
-      const btn = page.locator(submitButtonSelector).first();
-      if ((await btn.count()) > 0) {
-        try {
-          await btn.click({ timeout: 2000 });
-        } catch {
-          await page.keyboard.press("Enter");
-        }
-      } else {
-        await page.keyboard.press("Enter");
-      }
-    } else {
-      await page.keyboard.press("Enter");
-    }
-    const tCaptureStart = Date.now();
-    const response = signalPromise
-      ? await Promise.race([responsePromise, signalPromise]).catch(() => null)
-      : await responsePromise.catch(() => null);
-    if (signal && abortListener) {
-      signal.removeEventListener("abort", abortListener);
-    }
-    if (response) {
-      await waitWithSignal(Math.min(postSubmitWaitMs, 30000), signal);
-    } else {
-      await waitWithSignal(postSubmitWaitMs, signal);
-    }
-    const captureResponseMs = Date.now() - tCaptureStart;
-    const submitMs = captureResponseMs;
-
-    let status = 0;
-    let contentType: string | null = null;
-    let body: Buffer = Buffer.alloc(0);
-    if (response) {
-      const captured = await readPageResponseBody(response);
-      if (captured.body.length > MAX_RESPONSE_BYTES) {
-        body = Buffer.from(
-          JSON.stringify({
-            error: {
-              message: "Response too large",
-              type: "upstream_error",
-            },
-          })
-        );
-        status = 502;
-        contentType = "application/json";
-      } else {
-        body = captured.body as unknown as Buffer;
-        contentType = captured.headers["content-type"] || null;
-      }
-    }
-
-    return {
-      status,
-      contentType,
-      body,
-      isStealth: pooled.isStealth,
-      timing: {
-        acquireContextMs,
-        navigateMs,
-        submitMs,
-        captureResponseMs,
-        totalMs: Date.now() - t0,
-      },
-    };
-  } catch (err) {
-    const msg = err instanceof Error ? err.message : String(err);
-    const body = Buffer.from(
-      JSON.stringify({
-        error: {
-          message: sanitizeErrorMessage(`browserBackedChat failed: ${msg}`),
-          type: "upstream_error",
-        },
-      })
-    );
-    return {
-      status: 502,
-      contentType: "application/json",
-      body,
-      isStealth: pooled.isStealth,
-      timing: {
-        acquireContextMs,
-        navigateMs: 0,
-        submitMs: 0,
-        captureResponseMs: 0,
-        totalMs: Date.now() - t0,
-      },
-    };
-  } finally {
-    await page.close();
-    if (!reuseAcquired) {
-      try {
-        await pooled.context.close();
-      } catch {
-        /* ignore */
-      }
-    }
-  }
-}

packages/browser-pool/src/services/browserPool.ts

@@ -1,440 +0,0 @@
-/**
- * browserPool.ts — Shared stealth browser pool for web-cookie providers.
- *
- * The DuckDuckGo VQD challenge and Claude web's Cloudflare Turnstile both
- * validate values that only a real browser can produce (DOM layout
- * measurements like offsetWidth/Height, getBoundingClientRect,
- * getComputedStyle, iframe contentWindow probes). Plain Node fetch + a
- * VM-stubs solver structurally runs the JS but cannot match those values,
- * so the server rejects the request.
- *
- * This pool keeps one Chromium instance warm and serves "browser contexts"
- * (one per provider) on demand. Each context owns one or more pages; the
- * caller is expected to be polite (one page per request, close on done).
- *
- * The pool prefers `cloakbrowser` (npm) when available — its binary-level
- * fingerprint patches (--fingerprint-timezone, --fingerprint-locale, and
- * dozens more) are the only thing that gets past DuckDuckGo's anti-bot
- * in this environment. Falls back to plain `playwright` if cloakbrowser
- * is not installed; the fallback works for Claude web (which only needs
- * valid cookies) but not for DDG's VQD challenge.
- *
- * Opt-in: pool only launches Chromium when an executor explicitly asks
- * for a context, so users who never use the browser-backed path pay zero
- * startup cost. Set OMNIROUTE_BROWSER_POOL=off to fully disable.
- */
-
-import { Buffer } from "node:buffer";
-import type {
-  BrowserPoolContextOptions,
-  BrowserPoolMetrics,
-  PooledContext,
-} from "../interfaces.ts";
-
-type Browser = import("playwright").Browser;
-type BrowserContext = import("playwright").BrowserContext;
-type Page = import("playwright").Page;
-
-/** Proxy resolver injected by the core stub after dynamic import. */
-type ProxyResolverFn = (
-  providerKey: string
-) => Promise<import("playwright").LaunchOptions["proxy"] | undefined>;
-
-let injectedProxyResolver: ProxyResolverFn | null = null;
-
-export function setProxyResolver(fn: ProxyResolverFn): void {
-  injectedProxyResolver = fn;
-}
-
-function createBrowserPoolMetrics(): BrowserPoolMetrics {
-  return {
-    browserLaunches: 0,
-    browserLaunchFailures: 0,
-    contextsCreated: 0,
-    contextsReused: 0,
-    contextsEvicted: 0,
-    contextsReleased: 0,
-    contextCreateFailures: 0,
-    shutdowns: 0,
-    lastShutdownReason: null,
-  };
-}
-
-interface PoolState {
-  browser: Browser | null;
-  contexts: Map<string, PooledContext>;
-  pendingContexts: Map<string, Promise<PooledContext>>;
-  launching: Promise<Browser> | null;
-  lastActivity: number;
-  idleTimer: NodeJS.Timeout | null;
-  evictTimer: NodeJS.Timeout | null;
-  cloakLaunch: ((opts: unknown) => Promise<Browser>) | null;
-  cloakLaunchResolved: boolean;
-  metrics: BrowserPoolMetrics;
-}
-
-const POOL_IDLE_TIMEOUT_MS = 5 * 60 * 1000;
-const CONTEXT_TTL_MS = 10 * 60 * 1000; // 10 min — evict stale contexts
-const EVICT_INTERVAL_MS = 60 * 1000; // check every 60s
-const DEFAULT_USER_AGENT =
-  "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/149.0.0.0 Safari/537.36";
-
-const state: PoolState = {
-  browser: null,
-  contexts: new Map(),
-  pendingContexts: new Map(),
-  launching: null,
-  lastActivity: 0,
-  idleTimer: null,
-  evictTimer: null,
-  cloakLaunch: null,
-  cloakLaunchResolved: false,
-  metrics: createBrowserPoolMetrics(),
-};
-
-function getCloakbrowserModuleId(): string {
-  // Keep this computed: cloakbrowser is an optional runtime enhancer, and a literal
-  // dynamic import with the package name makes Turbopack resolve it during route compilation.
-  return ["cloak", "browser"].join("");
-}
-
-async function resolveCloakLaunch(): Promise<((opts: unknown) => Promise<Browser>) | null> {
-  if (state.cloakLaunchResolved) return state.cloakLaunch;
-  state.cloakLaunchResolved = true;
-  try {
-    const mod = (await import(getCloakbrowserModuleId())) as unknown as {
-      launch?: (opts: unknown) => Promise<Browser>;
-    };
-    state.cloakLaunch = mod.launch ?? null;
-  } catch {
-    state.cloakLaunch = null;
-  }
-  return state.cloakLaunch;
-}
-
-function isPoolEnabled(): boolean {
-  const flag = process.env.OMNIROUTE_BROWSER_POOL;
-  if (flag === undefined) return true;
-  return flag !== "off" && flag !== "0" && flag !== "false";
-}
-
-function resetIdleTimer(): void {
-  if (state.idleTimer) clearTimeout(state.idleTimer);
-  state.idleTimer = setTimeout(() => {
-    void shutdownPool("idle-timeout");
-  }, POOL_IDLE_TIMEOUT_MS);
-  state.idleTimer.unref?.();
-}
-
-function evictStaleContexts(): void {
-  const now = Date.now();
-  for (const [key, pooled] of state.contexts) {
-    if (now - pooled.lastUsed > CONTEXT_TTL_MS) {
-      console.log(
-        "[BrowserPool] Evicted stale context:",
-        key,
-        "(idle",
-        ((now - pooled.lastUsed) / 1000).toFixed(0) + "s)"
-      );
-      state.contexts.delete(key);
-      state.metrics.contextsEvicted++;
-      pooled.context.close().catch(() => {});
-    }
-  }
-  if (state.contexts.size === 0 && !state.launching) {
-    void shutdownPool("all-contexts-evicted");
-  }
-}
-
-function startEvictTimer(): void {
-  if (state.evictTimer) clearInterval(state.evictTimer);
-  state.evictTimer = setInterval(() => evictStaleContexts(), EVICT_INTERVAL_MS);
-  state.evictTimer.unref?.();
-}
-
-async function launchBrowser(): Promise<Browser> {
-  if (state.browser) return state.browser;
-  if (state.launching) return state.launching;
-  state.launching = (async () => {
-    const cloakLaunch = await resolveCloakLaunch();
-    let browser: Browser;
-    if (cloakLaunch) {
-      browser = await cloakLaunch({
-        headless: true,
-        args: ["--no-sandbox", "--disable-dev-shm-usage"],
-      });
-    } else {
-      // Fallback: plain Playwright. Works for Claude web (cookie-only
-      // auth) but DDG's VQD challenge will detect this Chromium build.
-      const { chromium } = await import("playwright");
-      browser = await chromium.launch({
-        headless: true,
-        args: [
-          "--no-sandbox",
-          "--disable-dev-shm-usage",
-          "--disable-blink-features=AutomationControlled",
-        ],
-      });
-    }
-    state.browser = browser;
-    state.launching = null;
-    state.metrics.browserLaunches++;
-    return browser;
-  })();
-  try {
-    return await state.launching;
-  } catch (err) {
-    state.launching = null;
-    state.metrics.browserLaunchFailures++;
-    throw err;
-  }
-}
-
-function parseCookieString(
-  raw: string,
-  domain: string
-): Array<{
-  name: string;
-  value: string;
-  domain: string;
-  path: string;
-  expires: number;
-  httpOnly: boolean;
-  secure: boolean;
-  sameSite: "Lax" | "Strict" | "None";
-}> {
-  return raw
-    .split(";")
-    .map((p) => p.trim())
-    .filter(Boolean)
-    .map((pair) => {
-      const eq = pair.indexOf("=");
-      if (eq < 0) return null;
-      const name = pair.slice(0, eq).trim();
-      const value = pair.slice(eq + 1).trim();
-      if (!name || !value) return null;
-      return {
-        name,
-        value,
-        domain: domain.startsWith(".") ? domain : `.${domain}`,
-        path: "/",
-        expires: -1,
-        httpOnly: false,
-        secure: true,
-        sameSite: "Lax" as const,
-      };
-    })
-    .filter(Boolean) as Array<{
-    name: string;
-    value: string;
-    domain: string;
-    path: string;
-    expires: number;
-    httpOnly: boolean;
-    secure: boolean;
-    sameSite: "Lax" | "Strict" | "None";
-  }>;
-}
-
-// Clear a key from the pending-creation map once its promise settles, counting
-// failures. Kept as a leaf helper so acquireBrowserContext stays under the
-// function-length ceiling (#3368 PR7 metrics).
-function settlePendingContext(key: string, failed: boolean): void {
-  if (failed) state.metrics.contextCreateFailures++;
-  state.pendingContexts.delete(key);
-}
-
-export async function acquireBrowserContext(
-  key: string,
-  options: BrowserPoolContextOptions
-): Promise<PooledContext> {
-  if (!isPoolEnabled()) {
-    throw new Error(
-      "browserPool: OMNIROUTE_BROWSER_POOL=off — context requested but pool is disabled"
-    );
-  }
-  const existing = state.contexts.get(key);
-  if (existing) {
-    existing.lastUsed = Date.now();
-    state.lastActivity = Date.now();
-    state.metrics.contextsReused++;
-    resetIdleTimer();
-    return existing;
-  }
-
-  // Dedup concurrent creations for the same key
-  const pending = state.pendingContexts.get(key);
-  if (pending) return pending;
-
-  const createPromise = (async (): Promise<PooledContext> => {
-    const proxy = injectedProxyResolver ? await injectedProxyResolver(key) : undefined;
-    const [browser] = await Promise.all([launchBrowser()]);
-    const isStealth = state.cloakLaunch !== null;
-    const context = await browser.newContext({
-      userAgent: options.userAgent || DEFAULT_USER_AGENT,
-      locale: options.locale || "en-US",
-      timezoneId: options.timezone || "America/New_York",
-      viewport: { width: 1280, height: 800 },
-      ...(proxy ? { proxy } : {}),
-    });
-
-    if (options.cookieString) {
-      const cookies = parseCookieString(options.cookieString, options.cookieDomain);
-      if (cookies.length > 0) {
-        await context.addCookies(cookies);
-      }
-    }
-
-    let warmupPage: Page | null = null;
-    if (options.warmupUrl) {
-      try {
-        warmupPage = await context.newPage();
-        await warmupPage.goto(options.warmupUrl, {
-          waitUntil: "domcontentloaded",
-          timeout: 30000,
-        });
-        // Give the warmup a moment for the upstream's status/auth/country
-        // JSON endpoints to fire. Without this, the first chat request would
-        // pay the warmup cost on the hot path.
-        await new Promise((r) => setTimeout(r, 1500));
-      } catch (err) {
-        try {
-          await warmupPage?.close();
-        } catch {
-          /* ignore */
-        }
-        warmupPage = null;
-        void err;
-      }
-    }
-
-    // Guard: if shutdownPool() ran while we were creating this context,
-    // the browser we obtained is now closed. Close our temp context and
-    // throw so the caller knows to retry.
-    if (state.browser !== browser) {
-      await context.close().catch(() => {});
-      if (warmupPage) {
-        await warmupPage.close().catch(() => {});
-      }
-      throw new Error("Pool shut down during context creation");
-    }
-
-    const pooled: PooledContext = {
-      id: key,
-      context,
-      warmupPage,
-      lastUsed: Date.now(),
-      isStealth,
-    };
-    state.contexts.set(key, pooled);
-    state.metrics.contextsCreated++;
-    state.lastActivity = Date.now();
-    resetIdleTimer();
-    startEvictTimer();
-    return pooled;
-  })();
-
-  state.pendingContexts.set(key, createPromise);
-  createPromise
-    .then(() => settlePendingContext(key, false))
-    .catch(() => settlePendingContext(key, true));
-
-  return createPromise;
-}
-
-export async function openPage(pooled: PooledContext): Promise<Page> {
-  return pooled.context.newPage();
-}
-
-export async function releaseBrowserContext(key: string): Promise<void> {
-  const pooled = state.contexts.get(key);
-  if (!pooled) return;
-  state.contexts.delete(key);
-  state.metrics.contextsReleased++;
-  try {
-    await pooled.context.close();
-  } catch {
-    /* ignore */
-  }
-  if (state.contexts.size === 0) {
-    await shutdownPool("last-context-closed");
-  }
-}
-
-export async function shutdownPool(reason: string): Promise<void> {
-  state.metrics.shutdowns++;
-  state.metrics.lastShutdownReason = reason;
-  if (state.idleTimer) {
-    clearTimeout(state.idleTimer);
-    state.idleTimer = null;
-  }
-  if (state.evictTimer) {
-    clearInterval(state.evictTimer);
-    state.evictTimer = null;
-  }
-  state.pendingContexts.clear();
-  for (const [key, pooled] of state.contexts) {
-    try {
-      await pooled.context.close();
-    } catch {
-      /* ignore */
-    }
-    state.contexts.delete(key);
-  }
-  if (state.browser) {
-    try {
-      await state.browser.close();
-    } catch {
-      /* ignore */
-    }
-    state.browser = null;
-  }
-  state.lastActivity = Date.now();
-  // Avoid unused-parameter lint: log reason via debug if anyone hooks
-  // process.on('exit') and prints state.
-  void reason;
-}
-
-function getBrowserPoolStatus(): {
-  enabled: boolean;
-  contexts: number;
-  browserRunning: boolean;
-  stealthAvailable: boolean;
-  lastActivityAgoMs: number;
-} {
-  return {
-    enabled: isPoolEnabled(),
-    contexts: state.contexts.size,
-    browserRunning: state.browser !== null,
-    stealthAvailable: state.cloakLaunch !== null,
-    lastActivityAgoMs: state.lastActivity === 0 ? -1 : Date.now() - state.lastActivity,
-  };
-}
-
-/**
- * #3368 PR7 — browser-pool observability. Returns live status plus cumulative
- * lifecycle telemetry (launches, context create/reuse/evict/release counts,
- * failures, shutdowns). Surfaced via the omniroute_browser_pool_status MCP tool.
- */
-export function getBrowserPoolMetrics(): {
-  status: ReturnType<typeof getBrowserPoolStatus>;
-  metrics: BrowserPoolMetrics;
-} {
-  return { status: getBrowserPoolStatus(), metrics: { ...state.metrics } };
-}
-
-/** Test-only: reset cumulative metrics so assertions start from a clean slate. */
-export function __resetBrowserPoolMetricsForTest(): void {
-  state.metrics = createBrowserPoolMetrics();
-}
-
-export async function readPageResponseBody(
-  response: import("playwright").Response
-): Promise<{ status: number; headers: Record<string, string>; body: Buffer }> {
-  const headers: Record<string, string> = {};
-  for (const [name, value] of Object.entries(response.headers())) {
-    headers[name] = value;
-  }
-  const body = await response.body();
-  return { status: response.status(), headers, body: Buffer.from(body) };
-}

packages/browser-pool/src/services/grokClearance.ts

@@ -1,73 +0,0 @@
-/**
- * grokClearance.ts — gated browser-backed cf_clearance acquisition for
- * grok-web (#8019).
- *
- * grok.com sits behind Cloudflare Enterprise, which pins `cf_clearance` to
- * the client's IP+TLS+UA fingerprint. Pure cookie-replay from
- * `grokTlsClient.ts` (TLS-impersonating fetch) cannot forge a fresh
- * clearance from a datacenter egress that Cloudflare has already flagged —
- * only a real browser solving the challenge natively can mint one bound to
- * that egress's own fingerprint.
- *
- * This module reuses the EXISTING provider-agnostic browser pool
- * (`browserPool.ts`, already live for claude-web + duckduckgo-web) rather
- * than adding a new Turnstile solver — `claudeTurnstileSolver.ts` is
- * claude.ai-specific and does not apply here.
- *
- * Opt-in only: gated behind `OMNIROUTE_BROWSER_POOL` / `WEB_COOKIE_USE_BROWSER`
- * (the same env gate already used by claude-web.ts / duckduckgo-web.ts).
- * With the gate off, `acquireFreshGrokClearance` is never called — the
- * executor stays on the Step-1 `cloudflare_challenge` classification.
- */
-
-import { acquireBrowserContext } from "./browserPool.ts";
-import type { PooledContext } from "../interfaces.ts";
-
-const GROK_WARMUP_URL = "https://grok.com/";
-const GROK_COOKIE_DOMAIN = ".grok.com";
-const GROK_POOL_KEY = "grok-web";
-
-/**
- * Reads the same opt-in gate as claude-web/duckduckgo-web
- * (`WEB_COOKIE_USE_BROWSER` or `OMNIROUTE_BROWSER_POOL`). Off by default.
- */
-export function shouldUseGrokBrowserBacked(): boolean {
-  const flag = process.env.WEB_COOKIE_USE_BROWSER;
-  if (flag === "1" || flag === "true" || flag === "on") return true;
-  const poolFlag = process.env.OMNIROUTE_BROWSER_POOL;
-  return poolFlag === "on" || poolFlag === "1" || poolFlag === "true";
-}
-
-async function readCfClearanceFromContext(pooled: PooledContext): Promise<string | null> {
-  const cookies = await pooled.context.cookies(GROK_WARMUP_URL);
-  const match = cookies.find((c) => c.name === "cf_clearance");
-  return match?.value || null;
-}
-
-async function acquireViaPool(): Promise<string | null> {
-  try {
-    const pooled = await acquireBrowserContext(GROK_POOL_KEY, {
-      cookieDomain: GROK_COOKIE_DOMAIN,
-      cookieString: null,
-      warmupUrl: GROK_WARMUP_URL,
-    });
-    return await readCfClearanceFromContext(pooled);
-  } catch {
-    return null;
-  }
-}
-
-/**
- * Acquire a fresh `.grok.com` cf_clearance via the shared browser pool.
- * Never throws — resolves to `null` on any failure so callers can fall
- * through to the Cloudflare-challenge error rather than crash the request.
- */
-export async function acquireFreshGrokClearance(
-  signal?: AbortSignal | null
-): Promise<string | null> {
-  try {
-    return await acquireViaPool();
-  } catch {
-    return null;
-  }
-}

packages/browser-pool/tsconfig.json

@@ -1,17 +0,0 @@
-{
-  "compilerOptions": {
-    "target": "ES2022",
-    "module": "esnext",
-    "moduleResolution": "bundler",
-    "noEmit": true,
-    "allowImportingTsExtensions": true,
-    "resolveJsonModule": true,
-    "skipLibCheck": true,
-    "esModuleInterop": true,
-    "strict": false,
-    "lib": ["esnext"],
-    "types": ["node"],
-    "allowJs": false
-  },
-  "include": ["src/**/*.ts"]
-}

pnpm-workspace.yaml

@@ -1,5 +1,4 @@
 packages:
-  - "packages/*"
   - "open-sse"
 allowBuilds:
   "@parcel/watcher": true

public/providers/cheaperinference.svg

@@ -0,0 +1,7 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 64 64" role="img" aria-labelledby="title">
+  <title id="title">Cheaper Inference gateway mark</title>
+  <rect x="1" y="1" width="62" height="62" fill="#050807" stroke="#31f889" stroke-width="2" />
+  <path d="M8 21V8h13M43 56h13V43" fill="none" stroke="#c8d2cc" stroke-width="2" />
+  <path d="M31 20c-8 0-13 5-13 12s5 12 13 12" fill="none" stroke="#f2f5f3" stroke-linecap="square" stroke-width="7" />
+  <path d="m33 25 10 7-10 7" fill="none" stroke="#31f889" stroke-linecap="square" stroke-linejoin="miter" stroke-width="5" />
+</svg>

scripts/check/check-known-symbols.ts

@@ -9,13 +9,17 @@
 //       não resolve para um executor válido é um símbolo morto (roteia para fallback
 //       silencioso em vez de falhar).
 //
-//   (2) COMBO STRATEGIES — a cadeia de despacho `strategy === "..."` em
-//       open-sse/services/combo.ts DEVE tratar exatamente o conjunto canônico de
-//       ROUTING_STRATEGY_VALUES (src/shared/constants/routingStrategies.ts), exceto
-//       as estratégias-default implícitas documentadas em IMPLICIT_DEFAULT_STRATEGIES
-//       (estratégias canônicas sem NENHUMA referência `strategy === "..."`; caem no
-//       ordenamento padrão). Adicionar um valor canônico sem fiá-lo no despacho, ou
-//       fiar uma string de estratégia que não é canônica (inventada), falha aqui.
+//   (2) COMBO STRATEGIES — o despacho DEVE tratar exatamente o conjunto canônico de
+//       ROUTING_STRATEGY_VALUES ∪ INTERNAL_ROUTING_STRATEGY_VALUES
+//       (src/shared/constants/routingStrategies.ts), exceto as estratégias-default
+//       implícitas documentadas em IMPLICIT_DEFAULT_STRATEGIES (estratégias canônicas
+//       sem ramo de despacho próprio; caem no ordenamento padrão). Em vez de casar
+//       literais `strategy === "..."` por regex sobre a fonte, o conjunto tratado
+//       (handled) vem de uma enumeração em runtime importada de
+//       open-sse/services/combo/strategyDispatch.ts — o módulo que importa as funções
+//       reais de ordenação/despacho e lista quais estratégias elas implementam. Adicionar
+//       um valor canônico sem fiá-lo no despacho/e na enumeração, ou fiar uma string de
+//       estratégia que não é canônica (inventada), falha aqui.
 //
 //   (3) TRANSLATOR PAIRS — os pares from:to registrados em runtime no registry de
 //       tradutores (após bootstrap) são congelados em KNOWN_TRANSLATOR_PAIRS. Catraca:
@@ -483,21 +487,15 @@ async function main(): Promise<void> {
     ...(strategiesMod.ROUTING_STRATEGY_VALUES as readonly string[]),
     ...(strategiesMod.INTERNAL_ROUTING_STRATEGY_VALUES as readonly string[]),
   ];
-  // The combo dispatch was decomposed (Block J): the `strategy === "..."` branches
-  // now live across combo.ts + its strategy-ordering leaves, so scan all of them.
-  const comboDispatchFiles = [
-    "open-sse/services/combo.ts",
-    "open-sse/services/combo/applyStrategyOrdering.ts",
-    "open-sse/services/combo/resolveAutoStrategy.ts",
-    // #3501: the fusion/pipeline dispatch branches moved here with the prelude
-    // extraction; the `strategy === "..."` checks are unchanged, just relocated.
-    "open-sse/services/combo/dispatchPrelude.ts",
-    "open-sse/services/combo/targetResolution.ts",
-  ];
-  const comboSource = comboDispatchFiles
-    .map((rel) => readFileSync(resolvePath(REPO_ROOT, rel), "utf8"))
-    .join("\n");
-  const handled = extractHandledStrategies(comboSource);
+  // G1: the handled set comes from a runtime-imported dispatch registry that imports the
+  // actual strategy-ordering functions and enumerates which strategies they implement —
+  // NOT from regex-scanning `strategy === "..."` literals in source. The old regex broke
+  // when the dispatch was decomposed (Block J / #3501) and will break again when R0.3
+  // converts it to a registry; enumerating at runtime keeps the gate correct either way.
+  // Each entry in HANDLED_COMBO_STRATEGIES must stay in sync with a real dispatch branch.
+  const strategyDispatchMod =
+    await import("@omniroute/open-sse/services/combo/strategyDispatch.ts");
+  const handled = new Set(strategyDispatchMod.HANDLED_COMBO_STRATEGIES as readonly string[]);
 
   // Stale-enforcement (6A.3): IMPLICIT_DEFAULT_STRATEGIES is a suppression allowlist —
   // each entry exists ONLY to suppress a `canonicalNotHandled` violation (a canonical

scripts/quality/test-scoped.sh

@@ -0,0 +1,111 @@
+#!/usr/bin/env bash
+# test-scoped — run only unit tests impacted by your changes.
+#
+# Usage:
+#   npm run test:scoped              # tests for changes vs HEAD~1
+#   npm run test:scoped -- --staged  # tests for staged changes only
+#
+# This is the local DX companion to the CI TIA gate (#8084 D1). The CI version
+# builds a full import-graph impact map; for local dev we use a fast heuristic:
+# - Changed test files → run those directly
+# - Changed source files → run tests that share the file's directory/name prefix
+# - Hub files (tsconfig, package.json, etc.) → suggest full suite
+#
+# For the full TIA (import-graph based), use: npm run test:scoped:full
+# (requires a pre-built impact map via: node scripts/quality/build-test-impact-map.mjs)
+
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+REPO_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)"
+
+# ── 1. Determine changed files ───────────────────────────────────────────────
+if [[ "${1:-}" == "--staged" ]]; then
+  CHANGED=$(git -C "$REPO_ROOT" diff --name-only --diff-filter=ACMR --cached)
+else
+  CHANGED=$(git -C "$REPO_ROOT" diff --name-only --diff-filter=ACMR HEAD~1...HEAD 2>/dev/null || \
+            git -C "$REPO_ROOT" diff --name-only --diff-filter=ACMR)
+fi
+
+if [ -z "$CHANGED" ]; then
+  echo "[test:scoped] No changed files — nothing to test."
+  exit 0
+fi
+
+# ── 2. Classify changes ──────────────────────────────────────────────────────
+HUB_RE="(setupPolyfill|tsconfig|package\\.json|package-lock\\.json|\\.env|vitest\\.config|stryker\\.conf)"
+TEST_FILES=()
+SRC_FILES=()
+HIT_HUB=false
+
+while IFS= read -r f; do
+  [ -z "$f" ] && continue
+  if echo "$f" | grep -qE "$HUB_RE"; then
+    HIT_HUB=true
+  elif echo "$f" | grep -qE '^tests/unit/.*\.test\.(ts|mjs)$'; then
+    TEST_FILES+=("$f")
+  elif echo "$f" | grep -qE '^(src|open-sse)/'; then
+    SRC_FILES+=("$f")
+  fi
+done <<< "$CHANGED"
+
+# ── 3. Hub file changed → full suite ─────────────────────────────────────────
+if [ "$HIT_HUB" = true ]; then
+  echo "[test:scoped] Hub file changed — run full suite: npm run test:unit"
+  exit 1
+fi
+
+# ── 4. Collect tests to run ──────────────────────────────────────────────────
+RUN_TESTS=()
+
+# Direct test file changes always run
+for tf in "${TEST_FILES[@]}"; do
+  RUN_TESTS+=("$tf")
+done
+
+# For source files, try the impact map first; fall back to heuristic
+MAP_FILE="$REPO_ROOT/config/quality/test-impact-map.json"
+if [ ${#SRC_FILES[@]} -gt 0 ] && [ -f "$MAP_FILE" ]; then
+  # Use the TIA selection with the impact map
+  SEL=$(printf '%s\n' "${SRC_FILES[@]}" | node "$REPO_ROOT/scripts/quality/select-impacted-tests.mjs" 2>/dev/null || echo "__RUN_ALL__")
+  if echo "$SEL" | grep -q "__RUN_ALL__"; then
+    echo "[test:scoped] Unmapped source change — run full suite: npm run test:unit"
+    exit 1
+  fi
+  while IFS= read -r t; do
+    [ -n "$t" ] && RUN_TESTS+=("$t")
+  done <<< "$SEL"
+elif [ ${#SRC_FILES[@]} -gt 0 ]; then
+  # No impact map — heuristic: suggest building it
+  echo "[test:scoped] No impact map found. Build it with: node scripts/quality/build-test-impact-map.mjs"
+  echo "[test:scoped] Or run the full suite: npm run test:unit"
+  echo ""
+  echo "[test:scoped] Changed source files:"
+  printf '  %s\n' "${SRC_FILES[@]}"
+  if [ ${#TEST_FILES[@]} -gt 0 ]; then
+    echo "[test:scoped] Running changed test files only..."
+  else
+    exit 1
+  fi
+fi
+
+# Deduplicate
+IFS=$'\n' SORTED=($(printf '%s\n' "${RUN_TESTS[@]}" | sort -u)); unset IFS
+
+if [ ${#SORTED[@]} -eq 0 ]; then
+  echo "[test:scoped] No impacted tests — source changes don't map to any unit test."
+  exit 0
+fi
+
+echo "[test:scoped] Running ${#SORTED[@]} impacted test(s)..."
+
+# ── 5. Run selected tests ────────────────────────────────────────────────────
+cd "$REPO_ROOT"
+exec cross-env \
+  DISABLE_SQLITE_AUTO_BACKUP=true \
+  node --max-old-space-size=8192 \
+  --import tsx/esm \
+  --import ./open-sse/utils/setupPolyfill.ts \
+  --import ./tests/_setup/isolateDataDir.ts \
+  --test --test-force-exit --test-concurrency=4 \
+  "${SORTED[@]}"

src/app/(dashboard)/dashboard/BootstrapBanner.tsx

@@ -27,22 +27,17 @@ export default function BootstrapBanner() {
       <span className="text-amber-500 dark:text-amber-400 text-base shrink-0 mt-0.5">⚠️</span>
       <div className="flex-1 min-w-0">
         <p className="font-semibold text-amber-900 dark:text-amber-300">
-          Running in zero-config mode
+          {t("zeroConfigBannerTitle")}
         </p>
         <p className="mt-0.5 text-amber-800/80 dark:text-amber-200/80">
-          OmniRoute auto-generated secure encryption keys on first launch. They are persisted to{" "}
-          <code className="font-mono bg-amber-200/50 dark:bg-amber-500/20 px-1 rounded text-xs">
-            {dataDir}
-          </code>
-          . No action is required — your data is encrypted and safe. To use custom keys, add{" "}
-          <code className="font-mono bg-amber-200/50 dark:bg-amber-500/20 px-1 rounded text-xs">
-            JWT_SECRET
-          </code>{" "}
-          and{" "}
-          <code className="font-mono bg-amber-200/50 dark:bg-amber-500/20 px-1 rounded text-xs">
-            STORAGE_ENCRYPTION_KEY
-          </code>{" "}
-          to that file.
+          {t.rich("zeroConfigBannerBody", {
+            dataDir,
+            code: (chunks) => (
+              <code className="font-mono bg-amber-200/50 dark:bg-amber-500/20 px-1 rounded text-xs">
+                {chunks}
+              </code>
+            ),
+          })}
         </p>
       </div>
       <button

src/app/(dashboard)/dashboard/HomePageClient.tsx

@@ -144,31 +144,31 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
     const cleanLatest = latest.replace(/^v/, "");
     if (platform === "darwin") {
       return {
-        label: "Download DMG (macOS)",
+        label: t("downloadDmg"),
         url: `https://github.com/diegosouzapw/OmniRoute/releases/download/v${cleanLatest}/OmniRoute-${cleanLatest}.dmg`,
-        desc: `A new version of the OmniRoute desktop app is available. Please download and install the macOS DMG installer to update (current: v${versionInfo?.current || ""}).`,
+        desc: t("downloadDmgDescription", { version: versionInfo?.current || "" }),
       };
     }
     if (platform === "win32") {
       return {
-        label: "Download EXE (Windows)",
+        label: t("downloadExe"),
         url: `https://github.com/diegosouzapw/OmniRoute/releases/download/v${cleanLatest}/OmniRoute.Setup.${cleanLatest}.exe`,
-        desc: `A new version of the OmniRoute desktop app is available. Please download and install the Windows EXE installer to update (current: v${versionInfo?.current || ""}).`,
+        desc: t("downloadExeDescription", { version: versionInfo?.current || "" }),
       };
     }
     if (platform === "linux") {
       return {
-        label: "Download AppImage (Linux)",
+        label: t("downloadAppImage"),
         url: `https://github.com/diegosouzapw/OmniRoute/releases/download/v${cleanLatest}/OmniRoute-${cleanLatest}.AppImage`,
-        desc: `A new version of the OmniRoute desktop app is available. Please download the Linux AppImage package to update (current: v${versionInfo?.current || ""}).`,
+        desc: t("downloadAppImageDescription", { version: versionInfo?.current || "" }),
       };
     }
     return {
-      label: "Download Update",
+      label: t("downloadUpdate"),
       url: `https://github.com/diegosouzapw/OmniRoute/releases/tag/v${cleanLatest}`,
-      desc: `A new version of the OmniRoute desktop app is available. Please download the respective app format for your system to update (current: v${versionInfo?.current || ""}).`,
+      desc: t("downloadUpdateDescription", { version: versionInfo?.current || "" }),
     };
-  }, [platform, versionInfo?.latest, versionInfo?.current]);
+  }, [platform, t, versionInfo?.latest, versionInfo?.current]);
 
   // Electron internal auto-updater state and listeners
   const [electronUpdateStatus, setElectronUpdateStatus] = useState<{
@@ -539,29 +539,29 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
               {
                 step: "install",
                 status: "done",
-                message: message || `Queued update to v${targetVersion}.`,
+                message: message || t("updateQueued", { version: targetVersion }),
               },
               {
                 step: "rebuild",
                 status: "running",
-                message: "Docker image is rebuilding in the background.",
+                message: t("updateDockerRebuilding"),
               },
               {
                 step: "restart",
                 status: "pending",
-                message: "Waiting for OmniRoute to restart with the new version.",
+                message: t("updateWaitingRestart"),
               },
             ]
           : [
               {
                 step: "install",
                 status: "running",
-                message: message || `Installing v${targetVersion}.`,
+                message: message || t("updateInstalling", { version: targetVersion }),
               },
               {
                 step: "restart",
                 status: "pending",
-                message: "Waiting for OmniRoute to restart with the new version.",
+                message: t("updateWaitingRestart"),
               },
             ];
 
@@ -593,14 +593,14 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
               next = mergeUpdateStep(next, {
                 step: "complete",
                 status: "done",
-                message: `OmniRoute is now running v${targetVersion}.`,
+                message: t("updateRunning", { version: targetVersion }),
               });
 
               return next;
             });
             setUpdating(false);
             setUpdatePhase("done");
-            notify.success(`OmniRoute updated to v${targetVersion}.`);
+            notify.success(t("updateCompleted", { version: targetVersion }));
             await fetchData();
             return;
           }
@@ -611,20 +611,20 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
               next = mergeUpdateStep(next, {
                 step: "rebuild",
                 status: "running",
-                message: `Docker image is still rebuilding for v${targetVersion}.`,
+                message: t("updateDockerStillRebuilding", { version: targetVersion }),
               });
             } else {
               next = mergeUpdateStep(next, {
                 step: "install",
                 status: "running",
-                message: `Installing v${targetVersion} in the background.`,
+                message: t("updateInstallingBackground", { version: targetVersion }),
               });
             }
 
             next = mergeUpdateStep(next, {
               step: "restart",
               status: "pending",
-              message: `Waiting for OmniRoute to come back on v${targetVersion}.`,
+              message: t("updateWaitingVersion", { version: targetVersion }),
             });
 
             return next;
@@ -636,20 +636,20 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
               next = mergeUpdateStep(next, {
                 step: "rebuild",
                 status: "running",
-                message: "Docker rebuild is still in progress.",
+                message: t("updateDockerStillInProgress"),
               });
             } else {
               next = mergeUpdateStep(next, {
                 step: "install",
                 status: "running",
-                message: `Installing v${targetVersion} in the background.`,
+                message: t("updateInstallingBackground", { version: targetVersion }),
               });
             }
 
             next = mergeUpdateStep(next, {
               step: "restart",
               status: "running",
-              message: "Service restart in progress. Waiting for OmniRoute to come back online...",
+              message: t("updateRestarting"),
             });
 
             return next;
@@ -661,14 +661,14 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
         mergeUpdateStep(prev, {
           step: "error",
           status: "failed",
-          message: `Update started, but v${targetVersion} did not become available before timeout. Refresh the page or check server logs.`,
+          message: t("updateTimeout", { version: targetVersion }),
         })
       );
       setUpdating(false);
       setUpdatePhase("failed");
-      notify.error(`Update to v${targetVersion} timed out.`);
+      notify.error(t("updateTimedOut", { version: targetVersion }));
     },
-    [fetchData]
+    [fetchData, t]
   );
 
   const handleUpdate = async () => {
@@ -689,12 +689,12 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
           // Passing the raw object to notify.error() rendered it as a React child →
           // "Minified React error #31" crash ("Internal Server Error" screen), e.g. on
           // the 403 from the loopback-only /api/system/version. Extract the string.
-          notify.error(extractApiErrorMessage(data, "Failed to start update."));
+          notify.error(extractApiErrorMessage(data, t("updateStartFailed")));
           setUpdating(false);
           setUpdatePhase("idle");
           return;
         }
-        notify.success(data.message || "Update started.");
+        notify.success(data.message || t("updateStarted"));
         await pollBackgroundUpdate({
           channel: data.channel || "docker-compose",
           message: data.message || "",
@@ -705,7 +705,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
 
       // SSE stream — read progress events
       if (!res.body) {
-        notify.error("No response stream received.");
+        notify.error(t("noResponseStream"));
         setUpdating(false);
         setUpdatePhase("idle");
         return;
@@ -735,10 +735,10 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
             if (event.step === "complete") {
               setUpdatePhase("done");
               setUpdating(false);
-              notify.success(event.message || "Update complete!");
+              notify.success(event.message || t("updateComplete"));
             } else if (event.step === "error") {
               setUpdatePhase("failed");
-              notify.error(event.message || "Update failed.");
+              notify.error(event.message || t("updateFailed"));
               setUpdating(false);
             }
           } catch {
@@ -753,7 +753,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
         {
           step: "error",
           status: "failed",
-          message: "Network error — connection lost during update.",
+          message: t("updateNetworkError"),
         },
       ]);
       setUpdating(false);
@@ -769,11 +769,11 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
     return () => clearTimeout(timer);
   }, [updatePhase]);
   const stepLabels: Record<string, string> = {
-    install: "Install Package",
-    rebuild: "Rebuild Native Modules",
-    restart: "Restart Service",
-    complete: "Complete",
-    error: "Error",
+    install: t("stepInstallPackage"),
+    rebuild: t("stepRebuildNativeModules"),
+    restart: t("stepRestartService"),
+    complete: t("stepComplete"),
+    error: t("stepError"),
   };
   const showUpdateOverlay = updatePhase !== "idle";
 
@@ -801,17 +801,17 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
               <div>
                 <h3 className="text-lg font-bold">
                   {updatePhase === "done"
-                    ? "Update Complete!"
+                    ? t("updateCompleteTitle")
                     : updatePhase === "failed"
-                      ? "Update Failed"
-                      : "Updating OmniRoute..."}
+                      ? t("updateFailedTitle")
+                      : t("updatingTitle")}
                 </h3>
                 <p className="text-xs text-text-muted mt-0.5">
                   {updatePhase === "done"
-                    ? "The page will reload automatically in a few seconds."
+                    ? t("reloadNotice")
                     : updatePhase === "failed"
-                      ? "Please try again or update manually via the CLI."
-                      : "Do not close this page. The system will restart automatically."}
+                      ? t("retryNotice")
+                      : t("restartNotice")}
                 </p>
               </div>
             </div>
@@ -871,7 +871,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                 <div className="mt-1 px-3 py-2.5 rounded-lg border border-green-500/30 bg-green-500/5">
                   <p className="text-sm font-semibold text-green-500 flex items-center gap-2">
                     <span className="material-symbols-outlined text-[18px]">check_circle</span>
-                    {updateSteps.find((s) => s.step === "complete")?.message || "Update complete!"}
+                    {updateSteps.find((s) => s.step === "complete")?.message || t("updateComplete")}
                   </p>
                   <p className="text-xs text-text-muted mt-1">{t("reloadingPageAutomatically")}</p>
                 </div>
@@ -891,11 +891,11 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                     if (updatePhase === "done") globalThis.window.location.reload();
                   }}
                 >
-                  {updatePhase === "done" ? "Reload Now" : "Close"}
+                  {updatePhase === "done" ? t("reloadNow") : t("closeUpdate")}
                 </Button>
                 {updatePhase === "failed" && (
                   <Button size="sm" variant="secondary" fullWidth onClick={handleUpdate}>
-                    Retry
+                    {t("retryUpdate")}
                   </Button>
                 )}
               </div>
@@ -917,30 +917,32 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                 </span>
                 <div>
                   <p className="font-semibold text-sm">
-                    Update Available: v{versionInfo.latest} {isElectron && "(Desktop App)"}
+                    {t("updateAvailableTitle", {
+                      version: versionInfo.latest,
+                      desktop: isElectron ? ` ${t("desktopAppLabel")}` : "",
+                    })}
                   </p>
                   <p className="text-xs opacity-80 mt-0.5">
                     {isElectron ? (
                       <>
-                        {electronUpdateStatus.status === "checking" && "Checking for updates..."}
+                        {electronUpdateStatus.status === "checking" && t("checkingForUpdates")}
                         {electronUpdateStatus.status === "available" &&
-                          `Version v${versionInfo.latest} is available for download.`}
+                          t("versionAvailableForDownload", { version: versionInfo.latest })}
                         {electronUpdateStatus.status === "downloading" &&
-                          `Downloading update... ${electronUpdateStatus.percent || 0}% complete.`}
-                        {electronUpdateStatus.status === "downloaded" &&
-                          "Update downloaded successfully! Click Restart & Install to apply."}
+                          t("downloadingUpdate", { percent: electronUpdateStatus.percent || 0 })}
+                        {electronUpdateStatus.status === "downloaded" && t("updateDownloaded")}
                         {electronUpdateStatus.status === "error" &&
-                          `Auto-update failed: ${electronUpdateStatus.message || "Unknown error"}.`}
+                          t("autoUpdateFailed", {
+                            reason: electronUpdateStatus.message || t("unknownUpdateError"),
+                          })}
                         {(electronUpdateStatus.status === "idle" ||
                           electronUpdateStatus.status === "not-available") &&
-                          `Version v${versionInfo.latest} is available for the desktop app.`}
+                          t("versionAvailableDesktop", { version: versionInfo.latest })}
                       </>
                     ) : versionInfo.autoUpdateSupported ? (
-                      t("updateAvailableDesc") ||
-                      `You are currently using v${versionInfo.current}. Update to access the latest features and bug fixes.`
+                      t("updateAvailableDesc")
                     ) : (
-                      versionInfo.autoUpdateError ||
-                      "Manual update required for this installation type."
+                      versionInfo.autoUpdateError || t("manualUpdateRequired")
                     )}
                   </p>
                 </div>
@@ -954,7 +956,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                       onClick={() => globalThis.window.electronAPI?.downloadUpdate()}
                       className="font-semibold"
                     >
-                      Download Update
+                      {t("downloadUpdate")}
                     </Button>
                   )}
                   {electronUpdateStatus.status === "downloading" && (
@@ -973,7 +975,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                       onClick={() => globalThis.window.electronAPI?.installUpdate()}
                       className="font-semibold animate-pulse"
                     >
-                      Restart & Install
+                      {t("restartAndInstall")}
                     </Button>
                   )}
                   {(electronUpdateStatus.status === "error" ||
@@ -989,7 +991,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                       }}
                       className="font-semibold"
                     >
-                      Check for Update
+                      {t("checkForUpdate")}
                     </Button>
                   )}
                 </div>
@@ -1001,9 +1003,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                   className="ml-4 shrink-0 font-semibold"
                   title={versionInfo.autoUpdateError || ""}
                 >
-                  {versionInfo.autoUpdateSupported
-                    ? t("updateNow") || "Update Now"
-                    : "Manual Update"}
+                  {versionInfo.autoUpdateSupported ? t("updateNow") : t("manualUpdate")}
                 </Button>
               )}
             </div>
@@ -1015,9 +1015,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                 electronUpdateStatus.status === "available" ||
                 electronUpdateStatus.status === "not-available") && (
                 <div className="flex flex-col sm:flex-row sm:items-center justify-between border-t border-primary/20 mt-2 pt-3 gap-2">
-                  <p className="text-xs opacity-75">
-                    Or download the respective installer format directly:
-                  </p>
+                  <p className="text-xs opacity-75">{t("directDownloadHint")}</p>
                   <div className="flex gap-2">
                     <Button
                       size="sm"
@@ -1029,7 +1027,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                       }
                       className="font-semibold text-xs py-1"
                     >
-                      Release Notes
+                      {t("releaseNotes")}
                     </Button>
                     <Button
                       size="sm"
@@ -1067,7 +1065,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
                   rel="noopener noreferrer"
                   className="ml-4 inline-flex shrink-0 items-center gap-1.5 rounded-lg border border-border bg-bg px-4 py-2 text-xs font-semibold text-text-main transition-colors hover:border-primary/30 hover:text-primary"
                 >
-                  {versionInfo.news.linkLabel || "Ler Mais"}
+                  {versionInfo.news.linkLabel || t("readMore")}
                   <span className="material-symbols-outlined text-[14px]">arrow_forward</span>
                 </a>
               )}
@@ -1214,7 +1212,7 @@ function ProviderOverviewCard({
     item.errors > 0 ? "text-red-500" : item.connected > 0 ? "text-green-500" : "text-text-muted";
 
   const authTypeConfig = {
-    "no-auth": { color: "bg-stone-500", label: "No Auth" },
+    "no-auth": { color: "bg-stone-500", label: t("noAuthLabel") },
     free: { color: "bg-green-500", label: tc("free") },
     oauth: { color: "bg-blue-500", label: t("oauthLabel") },
     apikey: { color: "bg-amber-500", label: t("apiKeyLabel") },

src/app/(dashboard)/dashboard/analytics/AutoRoutingAnalyticsTab.tsx

@@ -42,9 +42,7 @@ export default function AutoRoutingAnalyticsTab() {
   if (!stats) {
     return (
       <Card>
-        <div className="text-center py-8 text-text-muted">
-          No auto-routing analytics available. Make requests using the auto/ prefix to see metrics.
-        </div>
+        <div className="text-center py-8 text-text-muted">{t("autoRoutingNoDataAvailable")}</div>
       </Card>
     );
   }
@@ -110,7 +108,9 @@ export default function AutoRoutingAnalyticsTab() {
             const percentage = stats.totalRequests > 0 ? (count / stats.totalRequests) * 100 : 0;
             return (
               <div key={variant} className="flex items-center gap-3">
-                <div className="w-32 text-sm font-medium capitalize">{variant || "default"}</div>
+                <div className="w-32 text-sm font-medium capitalize">
+                  {variant || t("defaultVariantLabel")}
+                </div>
                 <div className="flex-1 h-3 bg-border rounded-full overflow-hidden">
                   <div
                     className="h-full bg-indigo-500 rounded-full transition-all"
@@ -133,9 +133,9 @@ export default function AutoRoutingAnalyticsTab() {
           <table className="w-full text-sm">
             <thead>
               <tr className="border-b border-border">
-                <th className="text-left py-2 px-3 font-medium">Provider</th>
-                <th className="text-right py-2 px-3 font-medium">Requests</th>
-                <th className="text-right py-2 px-3 font-medium">Share</th>
+                <th className="text-left py-2 px-3 font-medium">{t("chartProvider")}</th>
+                <th className="text-right py-2 px-3 font-medium">{t("chartRequests")}</th>
+                <th className="text-right py-2 px-3 font-medium">{t("chartShare")}</th>
               </tr>
             </thead>
             <tbody>

src/app/(dashboard)/dashboard/analytics/SearchAnalyticsTab.tsx

@@ -50,11 +50,13 @@ function ProviderBar({
   count,
   total,
   costUsd,
+  queriesLabel,
 }: {
   provider: string;
   count: number;
   total: number;
   costUsd: number;
+  queriesLabel: string;
 }) {
   const pct = total > 0 ? Math.round((count / total) * 100) : 0;
   return (
@@ -62,7 +64,7 @@ function ProviderBar({
       <div className="flex justify-between text-sm">
         <span className="font-medium text-text">{provider}</span>
         <span className="text-text-muted">
-          {count} queries · ${costUsd.toFixed(4)}
+          {count} {queriesLabel} · ${costUsd.toFixed(4)}
         </span>
       </div>
       <div className="h-2 rounded-full bg-bg-muted overflow-hidden">
@@ -99,7 +101,7 @@ export default function SearchAnalyticsTab() {
     return (
       <div className="flex items-center justify-center py-16 text-text-muted">
         <span className="material-symbols-outlined animate-spin mr-2">progress_activity</span>
-        Loading search analytics…
+        {t("searchAnalyticsLoading")}
       </div>
     );
   }
@@ -108,9 +110,11 @@ export default function SearchAnalyticsTab() {
     return (
       <div className="card p-6 text-center text-text-muted">
         <span className="material-symbols-outlined text-[32px] mb-2 block">search_off</span>
-        {error || "No search data available yet."}
+        {error || t("searchAnalyticsNoData")}
         <p className="text-xs mt-2">
-          Search requests will appear here after the first search via /v1/search.
+          {t.rich("searchAnalyticsNoDataDescription", {
+            code: (chunks) => <code className="bg-bg-muted px-1 rounded">{chunks}</code>,
+          })}
         </p>
       </div>
     );
@@ -126,25 +130,29 @@ export default function SearchAnalyticsTab() {
           icon="manage_search"
           label={t("searchAnalyticsTotalSearches")}
           value={stats.total.toLocaleString()}
-          sub={`${stats.today} today`}
+          sub={t("searchAnalyticsToday", { count: stats.today })}
         />
         <StatCard
           icon="cached"
           label={t("searchAnalyticsCacheHitRate")}
           value={`${stats.cacheHitRate}%`}
-          sub={`${stats.cached} cached requests`}
+          sub={t("searchAnalyticsCachedRequests", { count: stats.cached })}
         />
         <StatCard
           icon="attach_money"
           label={t("searchAnalyticsTotalCost")}
           value={`$${stats.totalCostUsd.toFixed(4)}`}
-          sub="search API costs"
+          sub={t("searchAnalyticsApiCosts")}
         />
         <StatCard
           icon="timer"
           label={t("searchAnalyticsAvgResponse")}
           value={`${stats.avgDurationMs}ms`}
-          sub={stats.errors > 0 ? `${stats.errors} errors` : "No errors"}
+          sub={
+            stats.errors > 0
+              ? t("searchAnalyticsErrors", { count: stats.errors })
+              : t("searchAnalyticsNoErrors")
+          }
         />
       </div>
 
@@ -153,7 +161,7 @@ export default function SearchAnalyticsTab() {
         <div className="card p-5">
           <h3 className="font-semibold text-text mb-4 flex items-center gap-2">
             <span className="material-symbols-outlined text-primary text-[20px]">hub</span>
-            Provider Breakdown
+            {t("searchAnalyticsProviderBreakdown")}
           </h3>
           <div className="flex flex-col gap-4">
             {providers.map(([prov, data]) => (
@@ -163,6 +171,7 @@ export default function SearchAnalyticsTab() {
                 count={data.count}
                 total={stats.total}
                 costUsd={data.costUsd}
+                queriesLabel={t("searchAnalyticsQueries")}
               />
             ))}
           </div>
@@ -177,8 +186,9 @@ export default function SearchAnalyticsTab() {
           </span>
           <p className="font-medium text-text">{t("searchAnalyticsNoSearchesYet")}</p>
           <p className="text-sm mt-1">
-            Use <code className="bg-bg-muted px-1 rounded">POST /v1/search</code> to start routing
-            web searches.
+            {t.rich("searchAnalyticsEmptyDescription", {
+              code: (chunks) => <code className="bg-bg-muted px-1 rounded">{chunks}</code>,
+            })}
           </p>
         </div>
       )}
@@ -189,8 +199,9 @@ export default function SearchAnalyticsTab() {
           check_circle
         </span>
         <span>
-          <strong>Free tier available:</strong> Serper (2,500/mo), Brave (2,000/mo), Exa (1,000/mo),
-          Tavily (1,000/mo) — total 6,500+ free searches/month with automatic failover.
+          {t.rich("searchAnalyticsFreeTier", {
+            strong: (chunks) => <strong>{chunks}</strong>,
+          })}
         </span>
       </div>
     </div>

src/app/(dashboard)/dashboard/audit/McpAuditTab.tsx

@@ -122,23 +122,23 @@ export default function McpAuditTab() {
         <div className="grid grid-cols-2 gap-3 md:grid-cols-4">
           {[
             {
-              label: "Calls (24h)",
+              label: t("mcpMetricCalls24h"),
               value: stats.totalCalls.toLocaleString(),
               icon: "terminal",
             },
             {
-              label: "Success rate",
+              label: t("mcpMetricSuccessRate"),
               value: `${Math.round(stats.successRate * 100)}%`,
               icon: "check_circle",
               highlight: stats.successRate >= 0.9,
             },
             {
-              label: "Avg duration",
+              label: t("mcpMetricAvgDuration"),
               value: `${Math.round(stats.avgDurationMs)}ms`,
               icon: "timer",
             },
             {
-              label: "Top tool",
+              label: t("mcpMetricTopTool"),
               value: stats.topTools[0]?.tool ?? "—",
               icon: "star",
             },

src/app/(dashboard)/dashboard/batch/BatchDetailModal.tsx

@@ -4,7 +4,9 @@ import { useEffect } from "react";
 import { useTranslations } from "next-intl";
 import { useBatchActions } from "./components/useBatchActions";
 
-function relativeTime(ts: number): string {
+type BatchTranslator = ReturnType<typeof useTranslations>;
+
+function relativeTime(ts: number, t: BatchTranslator): string {
   const diffMs = Date.now() - ts * 1000;
   const isFuture = diffMs < 0;
   const absDiffMs = Math.abs(diffMs);
@@ -22,8 +24,9 @@ function relativeTime(ts: number): string {
     }
   }
 
-  if (isFuture) return `in ${res}`;
-  return `${res} ago`;
+  return isFuture
+    ? t("batchRelativeTimeIn", { value: res })
+    : t("batchRelativeTimeAgo", { value: res });
 }
 
 interface BatchRecord {
@@ -93,6 +96,22 @@ const STATUS_LABELS: Record<string, string> = {
   expired_with_failures: "expired (partial)",
 };
 
+const STATUS_TRANSLATION_KEYS: Record<string, string> = {
+  completed: "batchStatusCompleted",
+  completed_with_failures: "batchStatusCompletedWithFailures",
+  failed: "batchStatusFailed",
+  in_progress: "batchStatusInProgress",
+  in_progress_with_failures: "batchStatusInProgressWithFailures",
+  finalizing: "batchStatusFinalizing",
+  finalizing_with_failures: "batchStatusFinalizingWithFailures",
+  validating: "batchStatusValidating",
+  cancelling: "batchStatusCancelling",
+  cancelled: "batchStatusCancelled",
+  cancelled_with_failures: "batchStatusCancelledWithFailures",
+  expired: "batchStatusExpired",
+  expired_with_failures: "batchStatusExpiredWithFailures",
+};
+
 function effectiveStatus(batch: BatchRecord): string {
   const hasFailed = (batch.requestCountsFailed ?? 0) > 0;
   if (!hasFailed) return batch.status;
@@ -106,10 +125,12 @@ function effectiveStatus(batch: BatchRecord): string {
   return map[batch.status] ?? batch.status;
 }
 
-function StatusBadge({ batch }: { batch: BatchRecord }) {
+function StatusBadge({ batch, t }: { batch: BatchRecord; t: BatchTranslator }) {
   const key = effectiveStatus(batch);
   const cls = STATUS_STYLES[key] ?? "bg-gray-500/15 text-gray-400 border-gray-500/25";
-  const label = STATUS_LABELS[key] ?? key.replace(/_/g, " ");
+  const label = STATUS_TRANSLATION_KEYS[key]
+    ? t(STATUS_TRANSLATION_KEYS[key])
+    : (STATUS_LABELS[key] ?? key.replace(/_/g, " "));
   return (
     <span className={`inline-block px-2 py-0.5 rounded-md text-xs font-medium border ${cls}`}>
       {label}
@@ -141,12 +162,24 @@ function formatTs(ts: number | null | undefined): string {
   });
 }
 
-export default function BatchDetailModal({ batch, files, onClose, onActionDone }: BatchDetailModalProps) {
+export default function BatchDetailModal({
+  batch,
+  files,
+  onClose,
+  onActionDone,
+}: BatchDetailModalProps) {
   const t = useTranslations("common");
 
   // ── Action hook (F7) ─────────────────────────────────────────────────────────
-  const { cancelling, retrying, error: actionError, cancel, retry, downloadHrefOutput, downloadHrefErrors } =
-    useBatchActions({ onRefresh: onActionDone, t });
+  const {
+    cancelling,
+    retrying,
+    error: actionError,
+    cancel,
+    retry,
+    downloadHrefOutput,
+    downloadHrefErrors,
+  } = useBatchActions({ onRefresh: onActionDone, t });
 
   // ── Status flags ──────────────────────────────────────────────────────────────
   const isTerminal = ["completed", "failed", "cancelled", "expired"].includes(batch.status);
@@ -195,8 +228,11 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
               pending_actions
             </span>
             <div>
-              <h2 id="batch-detail-modal-title" className="text-base font-semibold text-[var(--color-text-main)]">
-                Batch Details
+              <h2
+                id="batch-detail-modal-title"
+                className="text-base font-semibold text-[var(--color-text-main)]"
+              >
+                {t("batchDetailsTitle")}
               </h2>
               <div className="flex items-center gap-2 mt-0.5">
                 <p className="text-xs text-[var(--color-text-muted)] font-mono">{batch.id}</p>
@@ -227,16 +263,18 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
           <div className="grid grid-cols-2 sm:grid-cols-3 gap-4">
             <div className="flex flex-col gap-0.5">
               <span className="text-[11px] uppercase tracking-wider font-medium text-[var(--color-text-muted)]">
-                Status
+                {t("status")}
               </span>
-              <StatusBadge batch={batch} />
+              <StatusBadge batch={batch} t={t} />
             </div>
             <Field label={t("batchDetailEndpoint")} value={batch.endpoint} />
             {batch.model && <Field label={t("batchDetailModel")} value={batch.model} />}
             <Field label={t("batchDetailWindow")} value={batch.completionWindow} />
             <Field
               label={t("batchDetailCreated")}
-              value={<span title={formatTs(batch.createdAt)}>{relativeTime(batch.createdAt)}</span>}
+              value={
+                <span title={formatTs(batch.createdAt)}>{relativeTime(batch.createdAt, t)}</span>
+              }
             />
           </div>
 
@@ -245,7 +283,7 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
             <div className="space-y-2">
               <div className="flex items-center justify-between text-xs">
                 <span className="text-[var(--color-text-muted)] uppercase tracking-wider font-medium">
-                  Progress
+                  {t("batchProgress")}
                 </span>
                 <span className="text-[var(--color-text-muted)]">
                   {completed} / {total} ({pct}%)
@@ -262,17 +300,9 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
                 />
               </div>
               <div className="flex gap-4 text-xs text-[var(--color-text-muted)]">
-                <span>
-                  <span className="text-emerald-400 font-medium">{completed}</span> completed
-                </span>
-                {failed > 0 && (
-                  <span>
-                    <span className="text-red-400 font-medium">{failed}</span> failed
-                  </span>
-                )}
-                <span>
-                  <span className="font-medium">{total - completed - failed}</span> pending
-                </span>
+                <span>{t("batchCompletedCount", { count: completed })}</span>
+                {failed > 0 && <span>{t("batchFailedCount", { count: failed })}</span>}
+                <span>{t("batchPendingCount", { count: total - completed - failed })}</span>
               </div>
             </div>
           )}
@@ -280,19 +310,19 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
           {/* Timestamps */}
           <div>
             <h3 className="text-[11px] uppercase tracking-wider font-medium text-[var(--color-text-muted)] mb-3">
-              Timeline
+              {t("batchTimeline")}
             </h3>
             <div className="grid grid-cols-2 sm:grid-cols-3 gap-3 text-sm">
               {[
-                { label: "Created", ts: batch.createdAt },
-                { label: "In Progress", ts: batch.inProgressAt },
-                { label: "Finalizing", ts: batch.finalizingAt },
-                { label: "Completed", ts: batch.completedAt },
-                { label: "Failed", ts: batch.failedAt },
-                { label: "Expires", ts: batch.expiresAt },
-                { label: "Expired", ts: batch.expiredAt },
-                { label: "Cancelling", ts: batch.cancellingAt },
-                { label: "Cancelled", ts: batch.cancelledAt },
+                { label: t("batchDetailCreated"), ts: batch.createdAt },
+                { label: t("batchTimelineInProgress"), ts: batch.inProgressAt },
+                { label: t("batchTimelineFinalizing"), ts: batch.finalizingAt },
+                { label: t("batchTimelineCompleted"), ts: batch.completedAt },
+                { label: t("batchTimelineFailed"), ts: batch.failedAt },
+                { label: t("batchTimelineExpires"), ts: batch.expiresAt },
+                { label: t("batchTimelineExpired"), ts: batch.expiredAt },
+                { label: t("batchTimelineCancelling"), ts: batch.cancellingAt },
+                { label: t("batchTimelineCancelled"), ts: batch.cancelledAt },
               ]
                 .filter((t) => t.ts)
                 .map(({ label, ts }) => (
@@ -315,9 +345,21 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
             </h3>
             <div className="space-y-2">
               {[
-                { role: "Input", fileId: batch.inputFileId, record: inputFile },
-                { role: "Output", fileId: batch.outputFileId, record: outputFile },
-                { role: "Errors", fileId: batch.errorFileId, record: errorFile },
+                {
+                  role: t("filesListUsedByRoleInput"),
+                  fileId: batch.inputFileId,
+                  record: inputFile,
+                },
+                {
+                  role: t("filesListUsedByRoleOutput"),
+                  fileId: batch.outputFileId,
+                  record: outputFile,
+                },
+                {
+                  role: t("filesListUsedByRoleError"),
+                  fileId: batch.errorFileId,
+                  record: errorFile,
+                },
               ]
                 .filter((f) => f.fileId)
                 .map(({ role, fileId, record }) => (
@@ -350,7 +392,7 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
                         className="flex items-center gap-1 px-2 py-1 text-xs rounded bg-[var(--color-surface)] border border-[var(--color-border)] text-[var(--color-text-muted)] hover:text-[var(--color-text-main)] transition-colors"
                       >
                         <span className="material-symbols-outlined text-[13px]">download</span>
-                        Download
+                        {t("filesListDownload")}
                       </a>
                     </div>
                   </div>
@@ -362,7 +404,7 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
           {batch.usage && (
             <div>
               <h3 className="text-[11px] uppercase tracking-wider font-medium text-[var(--color-text-muted)] mb-3">
-                Token Usage
+                {t("batchTokenUsage")}
               </h3>
               <pre className="p-3 rounded-lg bg-[var(--color-bg-alt)] border border-[var(--color-border)] text-xs font-mono text-[var(--color-text-main)] overflow-x-auto">
                 {JSON.stringify(batch.usage, null, 2)}
@@ -374,7 +416,7 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
           {batch.errors && (
             <div>
               <h3 className="text-[11px] uppercase tracking-wider font-medium text-red-400 mb-3">
-                Errors
+                {t("errors")}
               </h3>
               <pre className="p-3 rounded-lg bg-red-500/5 border border-red-500/20 text-xs font-mono text-red-300 overflow-x-auto">
                 {JSON.stringify(batch.errors, null, 2)}
@@ -386,7 +428,7 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
           {batch.metadata && Object.keys(batch.metadata).length > 0 && (
             <div>
               <h3 className="text-[11px] uppercase tracking-wider font-medium text-[var(--color-text-muted)] mb-3">
-                Metadata
+                {t("batchMetadata")}
               </h3>
               <div className="space-y-1">
                 {Object.entries(batch.metadata).map(([k, v]) => (
@@ -438,7 +480,7 @@ export default function BatchDetailModal({ batch, files, onClose, onActionDone }
                   if (
                     window.confirm(
                       t("batchDetailActionRetry") +
-                        ` (${batch.requestCountsFailed} ${t("batchActionRetry")})?`,
+                        ` (${batch.requestCountsFailed} ${t("batchActionRetry")})?`
                     )
                   ) {
                     const result = await retry({

src/app/(dashboard)/dashboard/batch/FileDetailModal.tsx

@@ -4,22 +4,24 @@ import { useEffect, useState } from "react";
 import { useTranslations } from "next-intl";
 import { Button } from "@/shared/components";
 
-function relativeTime(ts: number): string {
+type FileTranslator = ReturnType<typeof useTranslations>;
+
+function relativeTime(ts: number, t: FileTranslator): string {
   const diffMs = Date.now() - ts * 1000;
   const diffSec = Math.round(diffMs / 1000);
-  if (diffSec < 60) return `${diffSec}s ago`;
+  if (diffSec < 60) return t("batchRelativeTimeAgo", { value: `${diffSec}s` });
   const diffMin = Math.round(diffSec / 60);
-  if (diffMin < 60) return `${diffMin}m ago`;
+  if (diffMin < 60) return t("batchRelativeTimeAgo", { value: `${diffMin}m` });
   const diffHr = Math.round(diffMin / 60);
-  if (diffHr < 24) return `${diffHr}h ago`;
+  if (diffHr < 24) return t("batchRelativeTimeAgo", { value: `${diffHr}h` });
   const diffDays = Math.round(diffHr / 24);
-  return `${diffDays}d ago`;
+  return t("batchRelativeTimeAgo", { value: `${diffDays}d` });
 }
 
-function relativeExpiration(ts: number | null): string {
-  if (!ts) return "Never";
+function relativeExpiration(ts: number | null, t: FileTranslator): string {
+  if (!ts) return t("batchFilesNeverExpires");
   const diffMs = ts * 1000 - Date.now();
-  if (diffMs <= 0) return "Expired";
+  if (diffMs <= 0) return t("expirationBadgeExpired");
   const diffSec = Math.round(diffMs / 1000);
   if (diffSec < 60) return `${diffSec}s`;
   const diffMin = Math.round(diffSec / 60);
@@ -55,6 +57,22 @@ interface BatchRecord {
   model?: string | null;
 }
 
+const BATCH_STATUS_TRANSLATION_KEYS: Record<string, string> = {
+  completed: "batchStatusCompleted",
+  completed_with_failures: "batchStatusCompletedWithFailures",
+  failed: "batchStatusFailed",
+  in_progress: "batchStatusInProgress",
+  in_progress_with_failures: "batchStatusInProgressWithFailures",
+  finalizing: "batchStatusFinalizing",
+  finalizing_with_failures: "batchStatusFinalizingWithFailures",
+  validating: "batchStatusValidating",
+  cancelling: "batchStatusCancelling",
+  cancelled: "batchStatusCancelled",
+  cancelled_with_failures: "batchStatusCancelledWithFailures",
+  expired: "batchStatusExpired",
+  expired_with_failures: "batchStatusExpiredWithFailures",
+};
+
 interface FileDetailModalProps {
   file: FileRecord;
   contents: string | null;
@@ -133,7 +151,7 @@ export default function FileDetailModal({
             </span>
             <div>
               <h2 className="text-base font-semibold text-[var(--color-text-main)]">
-                File Contents
+                {t("batchFileContents")}
               </h2>
               <div className="flex items-center gap-2 mt-0.5">
                 <p className="text-xs text-[var(--color-text-muted)] font-mono">{file.id}</p>
@@ -164,7 +182,7 @@ export default function FileDetailModal({
           <div className="grid grid-cols-2 sm:grid-cols-4 gap-4 p-4 rounded-xl bg-[var(--color-bg-alt)] border border-[var(--color-border)]">
             <div className="flex flex-col gap-0.5">
               <span className="text-[10px] uppercase tracking-wider font-medium text-[var(--color-text-muted)]">
-                Size
+                {t("batchFilesSizeColumn")}
               </span>
               <span className="text-sm text-[var(--color-text-main)]">
                 {formatBytes(file.bytes)}
@@ -172,24 +190,24 @@ export default function FileDetailModal({
             </div>
             <div className="flex flex-col gap-0.5">
               <span className="text-[10px] uppercase tracking-wider font-medium text-[var(--color-text-muted)]">
-                Purpose
+                {t("batchFilesPurpose")}
               </span>
               <span className="text-sm text-[var(--color-text-main)]">{file.purpose}</span>
             </div>
             <div className="flex flex-col gap-0.5">
               <span className="text-[10px] uppercase tracking-wider font-medium text-[var(--color-text-muted)]">
-                Created
+                {t("batchDetailCreated")}
               </span>
               <span className="text-sm text-[var(--color-text-main)]">
-                {createdAtTs ? relativeTime(createdAtTs) : "—"}
+                {createdAtTs ? relativeTime(createdAtTs, t) : "—"}
               </span>
             </div>
             <div className="flex flex-col gap-0.5">
               <span className="text-[10px] uppercase tracking-wider font-medium text-[var(--color-text-muted)]">
-                Expires
+                {t("batchFilesExpires")}
               </span>
               <span className="text-sm text-[var(--color-text-main)]">
-                {expiresAtTs ? relativeExpiration(expiresAtTs) : "Never"}
+                {expiresAtTs ? relativeExpiration(expiresAtTs, t) : t("batchFilesNeverExpires")}
               </span>
             </div>
           </div>
@@ -198,7 +216,7 @@ export default function FileDetailModal({
           {relatedBatches.length > 0 && (
             <div>
               <h3 className="text-[11px] uppercase tracking-wider font-medium text-[var(--color-text-muted)] mb-2">
-                Used by {relatedBatches.length} batch{relatedBatches.length > 1 ? "es" : ""}
+                {t("batchFileUsedByCount", { count: relatedBatches.length })}
               </h3>
               <div className="space-y-1.5">
                 {relatedBatches.map((b) => (
@@ -219,7 +237,9 @@ export default function FileDetailModal({
                             : "bg-gray-500/15 text-gray-400 border-gray-500/25"
                       }`}
                     >
-                      {b.status.replaceAll("_", " ")}
+                      {BATCH_STATUS_TRANSLATION_KEYS[b.status]
+                        ? t(BATCH_STATUS_TRANSLATION_KEYS[b.status])
+                        : b.status.replaceAll("_", " ")}
                     </span>
                   </div>
                 ))}
@@ -231,7 +251,7 @@ export default function FileDetailModal({
           <div className="flex-1 flex flex-col min-h-[300px]">
             <div className="flex items-center justify-between mb-2">
               <h3 className="text-xs font-semibold uppercase tracking-wider text-[var(--color-text-muted)]">
-                Preview
+                {t("batchFilePreview")}
               </h3>
               {contents && (
                 <button
@@ -241,7 +261,7 @@ export default function FileDetailModal({
                   <span className="material-symbols-outlined text-[14px]">
                     {copied ? "check" : "content_copy"}
                   </span>
-                  {copied ? "Copied!" : "Copy"}
+                  {copied ? t("copied") : t("copy")}
                 </button>
               )}
             </div>
@@ -259,7 +279,7 @@ export default function FileDetailModal({
                   {isTruncated && (
                     <div className="mt-3 p-3 bg-yellow-500/10 border border-yellow-500/25 rounded-lg text-xs text-yellow-400 flex items-center gap-2">
                       <span className="material-symbols-outlined text-[16px]">warning</span>
-                      Showing first 1000 lines ({lineCount} total lines)
+                      {t("batchFilePreviewTruncated", { shown: 1000, total: lineCount })}
                     </div>
                   )}
                 </div>
@@ -281,7 +301,7 @@ export default function FileDetailModal({
               className="flex items-center justify-center gap-2 px-4 py-2 text-sm font-medium rounded-lg bg-[var(--color-accent)] text-white hover:opacity-90 transition-opacity"
             >
               <span className="material-symbols-outlined text-[18px]">download</span>
-              Download Full File
+              {t("batchFileDownloadFull")}
             </Button>
           </div>
         </div>

src/app/(dashboard)/dashboard/chaos/chaosI18n.ts

@@ -0,0 +1,16 @@
+export type ChaosTranslator = ((
+  key: string,
+  values?: Record<string, string | number>
+) => string) & {
+  has?: (key: string) => boolean;
+};
+
+export function chaosText(
+  t: ChaosTranslator,
+  key: string,
+  fallback: string,
+  values?: Record<string, string | number>
+): string {
+  if (typeof t.has !== "function" || !t.has(key)) return fallback;
+  return values ? t(key, values) : t(key);
+}

src/app/(dashboard)/dashboard/chaos/components/ChaosConfigActionsBar.tsx

@@ -1,6 +1,7 @@
 "use client";
 
 import { useTranslations } from "next-intl";
+import { chaosText, type ChaosTranslator } from "../chaosI18n";
 
 /**
  * Save/Reset + Test-run action buttons for the Chaos Mode config page.
@@ -22,7 +23,7 @@ export function ChaosConfigActionsBar({
   onReset: () => void;
   onTest: () => void;
 }) {
-  const t = useTranslations("chaosConfig");
+  const t = useTranslations("chaosConfig") as ChaosTranslator;
 
   return (
     <>
@@ -66,7 +67,7 @@ export function ChaosConfigActionsBar({
           ) : (
             <span className="material-symbols-outlined text-[16px]">play_arrow</span>
           )}
-          {testing ? "Running..." : t("testButton")}
+          {testing ? chaosText(t, "running", "Running...") : t("testButton")}
         </button>
       </div>
     </>

src/app/(dashboard)/dashboard/chaos/components/ChaosTestResultsPanel.tsx

@@ -1,5 +1,8 @@
 "use client";
 
+import { useTranslations } from "next-intl";
+import { chaosText, type ChaosTranslator } from "../chaosI18n";
+
 export interface ChaosModelResult {
   providerId: string;
   providerName: string;
@@ -26,14 +29,24 @@ export interface ChaosTestResult {
  * complexity/size ratchet (config/quality/complexity-baseline.json).
  */
 export function ChaosTestResultsPanel({ result }: { result: ChaosTestResult }) {
+  const t = useTranslations("chaosConfig") as ChaosTranslator;
+  const resultsTitle = chaosText(
+    t,
+    "testResults",
+    `Test Results — ${result.mode} mode (${result.totalProviders} providers)`,
+    { mode: result.mode, count: result.totalProviders }
+  );
+  const startedLabel = chaosText(
+    t,
+    "started",
+    `Started: ${new Date(result.startedAt).toLocaleTimeString()}`,
+    { time: new Date(result.startedAt).toLocaleTimeString() }
+  );
+
   return (
     <div className="p-3 rounded-lg border border-border bg-surface/40 space-y-3">
-      <h3 className="text-sm font-bold text-text-main">
-        Test Results — {result.mode} mode ({result.totalProviders} providers)
-      </h3>
-      <div className="text-xs text-text-muted">
-        Started: {new Date(result.startedAt).toLocaleTimeString()}
-      </div>
+      <h3 className="text-sm font-bold text-text-main">{resultsTitle}</h3>
+      <div className="text-xs text-text-muted">{startedLabel}</div>
       {result.models.map((model, idx) => (
         <div
           key={idx}

src/app/(dashboard)/dashboard/chaos/page.tsx

@@ -1,11 +1,13 @@
 /**
  * /dashboard/chaos/page.tsx — Chaos Mode Configuration
  */
+import { getTranslations } from "next-intl/server";
 import ChaosConfigPageClient from "./ChaosConfigPageClient";
 
-export const metadata = {
-  title: "Chaos Mode — OmniRoute",
-};
+export async function generateMetadata() {
+  const t = await getTranslations("chaosConfig");
+  return { title: `${t("pageTitle")} — OmniRoute` };
+}
 
 export default function Page() {
   return <ChaosConfigPageClient />;

src/app/(dashboard)/dashboard/chaos/useChaosConfigPersistence.ts

@@ -2,6 +2,7 @@
 
 import { useCallback, useState, type Dispatch, type SetStateAction } from "react";
 import { useTranslations } from "next-intl";
+import { chaosText, type ChaosTranslator } from "./chaosI18n";
 import type { ChaosPageConfig, ChaosPageMessage } from "./chaosPageTypes";
 
 /**
@@ -13,7 +14,7 @@ export function useChaosConfigPersistence(
   config: ChaosPageConfig,
   setConfig: Dispatch<SetStateAction<ChaosPageConfig>>
 ) {
-  const t = useTranslations("chaosConfig");
+  const t = useTranslations("chaosConfig") as ChaosTranslator;
   const [saving, setSaving] = useState(false);
   const [message, setMessage] = useState<ChaosPageMessage>(null);
 
@@ -49,17 +50,26 @@ export function useChaosConfigPersistence(
       if (res.ok) {
         const data = await res.json();
         setConfig(data.config);
-        setMessage({ type: "success", text: "Config reset to defaults" });
+        setMessage({
+          type: "success",
+          text: chaosText(t, "resetSuccess", "Config reset to defaults"),
+        });
       } else {
-        const err = await res.json().catch(() => ({ error: "Reset failed" }));
-        setMessage({ type: "error", text: err.error || "Reset failed" });
+        const err = await res.json().catch(() => ({ error: null }));
+        setMessage({
+          type: "error",
+          text: err.error || chaosText(t, "resetFailed", "Reset failed"),
+        });
       }
     } catch {
-      setMessage({ type: "error", text: "Failed to reset config" });
+      setMessage({
+        type: "error",
+        text: chaosText(t, "resetFailed", "Failed to reset config"),
+      });
     } finally {
       setSaving(false);
     }
-  }, [setConfig]);
+  }, [setConfig, t]);
 
   return { t, saving, message, setMessage, saveConfig, resetConfig };
 }

src/app/(dashboard)/dashboard/chaos/useChaosTestRun.ts

@@ -2,6 +2,7 @@
 
 import { useCallback, useState } from "react";
 import { useTranslations } from "next-intl";
+import { chaosText, type ChaosTranslator } from "./chaosI18n";
 import type { ChaosTestResult } from "./components/ChaosTestResultsPanel";
 import type { ChaosPageConfig, ChaosPageMessage } from "./chaosPageTypes";
 
@@ -10,8 +11,11 @@ import type { ChaosPageConfig, ChaosPageMessage } from "./chaosPageTypes";
  * of the page component to keep it under the complexity/size ratchet
  * (config/quality/complexity-baseline.json).
  */
-export function useChaosTestRun(config: ChaosPageConfig, setMessage: (message: ChaosPageMessage) => void) {
-  const t = useTranslations("chaosConfig");
+export function useChaosTestRun(
+  config: ChaosPageConfig,
+  setMessage: (message: ChaosPageMessage) => void
+) {
+  const t = useTranslations("chaosConfig") as ChaosTranslator;
   const [testing, setTesting] = useState(false);
   const [testResult, setTestResult] = useState<ChaosTestResult | null>(null);
 
@@ -34,11 +38,17 @@ export function useChaosTestRun(config: ChaosPageConfig, setMessage: (message: C
         const data: ChaosTestResult = await res.json();
         setTestResult(data);
       } else {
-        const err = await res.json().catch(() => ({ error: "Unknown error" }));
-        setMessage({ type: "error", text: err.error || "Test failed" });
+        const err = await res.json().catch(() => ({ error: null }));
+        setMessage({
+          type: "error",
+          text: err.error || chaosText(t, "testFailed", "Test failed"),
+        });
       }
     } catch (err: any) {
-      setMessage({ type: "error", text: err.message || "Test failed" });
+      setMessage({
+        type: "error",
+        text: err.message || chaosText(t, "testFailed", "Test failed"),
+      });
     } finally {
       setTesting(false);
     }

src/app/(dashboard)/dashboard/cli-code/components/CliproxyapiToolCard.tsx

@@ -1,6 +1,7 @@
 "use client";
 
 import { useState, useEffect, useCallback } from "react";
+import { useTranslations } from "next-intl";
 import { Card, Button } from "@/shared/components";
 
 interface ToolState {
@@ -24,6 +25,7 @@ interface UpdateInfo {
 }
 
 export default function CliproxyapiToolCard({ isExpanded = false, onToggle = () => {} }) {
+  const t = useTranslations("cliTools");
   const [toolState, setToolState] = useState<ToolState | null>(null);
   const [updateInfo, setUpdateInfo] = useState<UpdateInfo | null>(null);
   const [loading, setLoading] = useState<string | null>(null);
@@ -71,7 +73,7 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
       });
       const data = await res.json();
       if (res.ok) {
-        setMessage({ type: "success", text: data.message || `${action} succeeded` });
+        setMessage({ type: "success", text: data.message || t("cliproxyapiActionSucceeded") });
         await fetchStatus();
         if (action === "install" || action === "restart") await fetchUpdateInfo();
       } else {
@@ -79,11 +81,14 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
           type: "error",
           text:
             (typeof data.error === "string" ? data.error : data.error?.message) ||
-            `${action} failed`,
+            t("cliproxyapiActionFailed"),
         });
       }
     } catch (err) {
-      setMessage({ type: "error", text: err instanceof Error ? err.message : "Request failed" });
+      setMessage({
+        type: "error",
+        text: err instanceof Error ? err.message : t("cliproxyapiRequestFailed"),
+      });
     } finally {
       setLoading(null);
     }
@@ -93,14 +98,26 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
     if (!toolState) return null;
     const s = toolState.status;
     const map: Record<string, { label: string; color: string }> = {
-      running: { label: "Running", color: "bg-green-500/10 text-green-600 dark:text-green-400" },
-      stopped: { label: "Stopped", color: "bg-zinc-500/10 text-zinc-500 dark:text-zinc-400" },
+      running: {
+        label: t("cliproxyapiStatusRunning"),
+        color: "bg-green-500/10 text-green-600 dark:text-green-400",
+      },
+      stopped: {
+        label: t("cliproxyapiStatusStopped"),
+        color: "bg-zinc-500/10 text-zinc-500 dark:text-zinc-400",
+      },
       not_installed: {
-        label: "Not Installed",
+        label: t("cliproxyapiStatusNotInstalled"),
         color: "bg-yellow-500/10 text-yellow-600 dark:text-yellow-400",
       },
-      installed: { label: "Installed", color: "bg-blue-500/10 text-blue-600 dark:text-blue-400" },
-      error: { label: "Error", color: "bg-red-500/10 text-red-600 dark:text-red-400" },
+      installed: {
+        label: t("cliproxyapiStatusInstalled"),
+        color: "bg-blue-500/10 text-blue-600 dark:text-blue-400",
+      },
+      error: {
+        label: t("cliproxyapiStatusError"),
+        color: "bg-red-500/10 text-red-600 dark:text-red-400",
+      },
     };
     const badge = map[s] || map.not_installed;
     return (
@@ -125,9 +142,7 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
               <h3 className="font-medium text-sm">CLIProxyAPI</h3>
               {statusBadge()}
             </div>
-            <p className="text-xs text-text-muted truncate">
-              Upstream proxy fallback (Go-based OAuth)
-            </p>
+            <p className="text-xs text-text-muted truncate">{t("cliproxyapiDescription")}</p>
           </div>
         </div>
         <span
@@ -161,7 +176,10 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
                   system_update
                 </span>
                 <span className="text-sm text-yellow-700 dark:text-yellow-300">
-                  Update available: v{updateInfo.current} → v{updateInfo.latest}
+                  {t("cliproxyapiUpdateAvailable", {
+                    current: updateInfo.current,
+                    latest: updateInfo.latest,
+                  })}
                 </span>
               </div>
               <Button
@@ -170,32 +188,34 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
                 onClick={() => apiCall("install", { version: updateInfo.latest })}
                 loading={loading === "install"}
               >
-                Update
+                {t("cliproxyapiUpdate")}
               </Button>
             </div>
           )}
 
           <div className="grid grid-cols-3 gap-3">
             <div className="p-3 rounded-lg bg-bg-secondary">
-              <p className="text-xs text-text-muted mb-1">Version</p>
+              <p className="text-xs text-text-muted mb-1">{t("cliproxyapiVersion")}</p>
               <p className="text-sm font-medium">
-                {toolState?.installedVersion ? `v${toolState.installedVersion}` : "Not installed"}
+                {toolState?.installedVersion
+                  ? `v${toolState.installedVersion}`
+                  : t("cliproxyapiNotInstalledValue")}
               </p>
             </div>
             <div className="p-3 rounded-lg bg-bg-secondary">
-              <p className="text-xs text-text-muted mb-1">Health</p>
+              <p className="text-xs text-text-muted mb-1">{t("cliproxyapiHealth")}</p>
               <p
                 className={`text-sm font-medium ${toolState?.healthStatus === "healthy" ? "text-green-600 dark:text-green-400" : toolState?.healthStatus === "unhealthy" ? "text-red-600 dark:text-red-400" : "text-text-muted"}`}
               >
                 {toolState?.healthStatus === "healthy"
-                  ? `Healthy`
+                  ? t("cliproxyapiHealthy")
                   : toolState?.healthStatus === "unhealthy"
-                    ? "Unhealthy"
-                    : "Unknown"}
+                    ? t("cliproxyapiUnhealthy")
+                    : t("cliproxyapiUnknown")}
               </p>
             </div>
             <div className="p-3 rounded-lg bg-bg-secondary">
-              <p className="text-xs text-text-muted mb-1">Port</p>
+              <p className="text-xs text-text-muted mb-1">{t("cliproxyapiPort")}</p>
               <p className="text-sm font-mono">{toolState?.port || 8317}</p>
             </div>
           </div>
@@ -209,7 +229,7 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
                 loading={loading === "install"}
               >
                 <span className="material-symbols-outlined text-[14px] mr-1">download</span>
-                Install
+                {t("cliproxyapiInstall")}
               </Button>
             )}
             {toolState?.status === "running" ? (
@@ -220,7 +240,7 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
                 loading={loading === "stop"}
               >
                 <span className="material-symbols-outlined text-[14px] mr-1">stop</span>
-                Stop
+                {t("cliproxyapiStop")}
               </Button>
             ) : toolState?.installedVersion ? (
               <Button
@@ -230,7 +250,7 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
                 loading={loading === "start"}
               >
                 <span className="material-symbols-outlined text-[14px] mr-1">play_arrow</span>
-                Start
+                {t("cliproxyapiStart")}
               </Button>
             ) : null}
             {toolState?.status === "running" && (
@@ -241,7 +261,7 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
                 loading={loading === "restart"}
               >
                 <span className="material-symbols-outlined text-[14px] mr-1">restart_alt</span>
-                Restart
+                {t("cliproxyapiRestart")}
               </Button>
             )}
             <Button
@@ -251,7 +271,7 @@ export default function CliproxyapiToolCard({ isExpanded = false, onToggle = ()
               loading={loading === "check"}
             >
               <span className="material-symbols-outlined text-[14px] mr-1">sync</span>
-              Check Updates
+              {t("cliproxyapiCheckUpdates")}
             </Button>
           </div>
         </div>

src/app/(dashboard)/dashboard/combos/IntelligentComboPanel.tsx

@@ -133,8 +133,8 @@ export default function IntelligentComboPanel({
               <code className="rounded bg-black/5 dark:bg-white/5 px-2 py-1 text-text-main">
                 {combo?.name}
               </code>
-              <span>{allCombos.length} intelligent combo(s)</span>
-              <span>{providerScopeCount} providers in scope</span>
+              <span>{t("intelligentComboCount", { count: allCombos.length })}</span>
+              <span>{t("providersInScope", { count: providerScopeCount })}</span>
             </div>
           </div>
 
@@ -161,7 +161,7 @@ export default function IntelligentComboPanel({
               </div>
               <div className="rounded-lg bg-black/5 dark:bg-white/5 px-3 py-2 text-right">
                 <p className="text-[10px] uppercase tracking-wide text-text-muted">
-                  Candidate Pool
+                  {getI18nOrFallback(t, "candidatePoolLabel", "Candidate Pool")}
                 </p>
                 <p className="text-lg font-semibold text-text-main">{providerScopeCount}</p>
               </div>
@@ -183,7 +183,12 @@ export default function IntelligentComboPanel({
                 </p>
               </div>
               {savingModePack && (
-                <span className="text-[11px] text-text-muted">Saving {savingModePack}…</span>
+                <span className="text-[11px] text-text-muted">
+                  {getI18nOrFallback(t, "savingModePack", "Saving {pack}…").replace(
+                    "{pack}",
+                    savingModePack
+                  )}
+                </span>
               )}
             </div>
 
@@ -308,7 +313,7 @@ export default function IntelligentComboPanel({
               </div>
               <div className="rounded-lg border border-black/8 bg-white/60 p-3 dark:border-white/8 dark:bg-white/[0.03]">
                 <p className="text-[11px] uppercase tracking-wide text-text-muted">
-                  Exploration Rate
+                  {getI18nOrFallback(t, "explorationRateLabel", "Exploration Rate")}
                 </p>
                 <p className="mt-1 text-sm font-semibold text-text-main">
                   {Math.round(normalizedConfig.explorationRate * 100)}%

src/app/(dashboard)/dashboard/combos/error.tsx

@@ -1,5 +1,7 @@
 "use client";
 
+import { useTranslations } from "next-intl";
+
 export default function CombosError({
   error: _error,
   reset,
@@ -7,6 +9,8 @@ export default function CombosError({
   error: Error & { digest?: string };
   reset: () => void;
 }) {
+  const t = useTranslations("combos");
+
   return (
     <div
       className="flex flex-col items-center justify-center min-h-[400px]"
@@ -14,14 +18,10 @@ export default function CombosError({
       aria-live="assertive"
     >
       <div className="text-center space-y-4">
-        <h2 className="text-xl font-semibold text-red-600 dark:text-red-400">
-          Failed to load combos
-        </h2>
-        <p className="text-text-muted max-w-md">
-          We could not load combo data right now. Check your connection and try again.
-        </p>
+        <h2 className="text-xl font-semibold text-red-600 dark:text-red-400">{t("errorTitle")}</h2>
+        <p className="text-text-muted max-w-md">{t("errorDescription")}</p>
         {_error?.digest && (
-          <p className="text-xs text-text-muted font-mono">Error ID: {_error.digest}</p>
+          <p className="text-xs text-text-muted font-mono">{t("errorId", { id: _error.digest })}</p>
         )}
         {process.env.NODE_ENV === "development" && _error?.message && (
           <p className="text-xs text-red-600 dark:text-red-400 font-mono">{_error.message}</p>
@@ -30,7 +30,7 @@ export default function CombosError({
           onClick={reset}
           className="px-4 py-2 bg-primary text-white rounded-lg hover:bg-primary-hover transition-colors focus:outline-2 focus:outline-offset-2 focus:outline-primary"
         >
-          Try Again
+          {t("errorRetry")}
         </button>
       </div>
     </div>

src/app/(dashboard)/dashboard/compression/page.tsx

@@ -1,9 +1,13 @@
 import { redirect } from "next/navigation";
+import { getTranslations } from "next-intl/server";
 
-export const metadata = {
-  title: "Compression",
-  description: "Configure context compression settings to reduce token usage and costs.",
-};
+export async function generateMetadata() {
+  const t = await getTranslations("metadata");
+  return {
+    title: t("compressionTitle"),
+    description: t("compressionDescription"),
+  };
+}
 
 export default function CompressionPage() {
   redirect("/dashboard/context/caveman");

src/app/(dashboard)/dashboard/endpoint/EndpointPageClient.tsx

@@ -534,7 +534,10 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
       if (!cloudConfigured) {
         setCloudStatus({
           type: "warning",
-          message: "Cloud sync is not configured on this instance.",
+          message: translateOrFallback(
+            "cloudSyncNotConfigured",
+            "Cloud sync is not configured on this instance."
+          ),
         });
         return;
       }
@@ -1822,7 +1825,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
           <div className="flex items-center gap-2 mb-3">
             <span className="material-symbols-outlined text-sm text-primary">hub</span>
             <h3 className="text-xs font-semibold text-text-muted uppercase tracking-wider">
-              {t("categoryCore") || "Core APIs"}
+              {t("categoryCore")}
             </h3>
             <div className="flex-1 h-px bg-border/50" />
           </div>
@@ -1843,7 +1846,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="code"
               iconColor="text-indigo-500"
               iconBg="bg-indigo-500/10"
-              title={t("responses") || "Responses API"}
+              title={t("responses")}
               path="/v1/responses"
               models={endpointData.chat}
               copy={copy}
@@ -1855,7 +1858,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="text_fields"
               iconColor="text-orange-500"
               iconBg="bg-orange-500/10"
-              title={t("completionsLegacy") || "Completions (Legacy)"}
+              title={t("completionsLegacy")}
               path="/v1/completions"
               models={endpointData.chat}
               copy={copy}
@@ -1867,7 +1870,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="psychology"
               iconColor="text-violet-500"
               iconBg="bg-violet-500/10"
-              title={t("messagesApi") || "Messages"}
+              title={t("messagesApi")}
               path="/v1/messages"
               models={null}
               badge="Anthropic"
@@ -1883,7 +1886,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
           <div className="flex items-center gap-2 mb-3">
             <span className="material-symbols-outlined text-sm text-purple-400">perm_media</span>
             <h3 className="text-xs font-semibold text-text-muted uppercase tracking-wider">
-              {t("categoryMedia") || "Media & Multi-Modal"}
+              {t("categoryMedia")}
             </h3>
             <div className="flex-1 h-px bg-border/50" />
           </div>
@@ -1916,7 +1919,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="edit_square"
               iconColor="text-violet-500"
               iconBg="bg-violet-500/10"
-              title={t("imageEdits") || "Image Edits"}
+              title={t("imageEdits")}
               path="/v1/images/edits"
               models={endpointData.images}
               copy={copy}
@@ -1952,7 +1955,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="music_note"
               iconColor="text-fuchsia-500"
               iconBg="bg-fuchsia-500/10"
-              title={t("musicGeneration") || "Music Generation"}
+              title={t("musicGeneration")}
               path="/v1/music/generations"
               models={endpointData.music}
               copy={copy}
@@ -1964,7 +1967,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="videocam"
               iconColor="text-red-500"
               iconBg="bg-red-500/10"
-              title={t("videoGeneration") || "Video Generation"}
+              title={t("videoGeneration")}
               path="/v1/videos/generations"
               models={endpointData.video}
               copy={copy}
@@ -1983,7 +1986,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
                 travel_explore
               </span>
               <h3 className="text-xs font-semibold text-text-muted uppercase tracking-wider">
-                {t("categorySearch") || "Search & Discovery"}
+                {t("categorySearch")}
               </h3>
               <div className="flex-1 h-px bg-border/50" />
             </div>
@@ -1992,7 +1995,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
                 icon="search"
                 iconColor="text-cyan-500"
                 iconBg="bg-cyan-500/10"
-                title={t("webSearch") || "Web Search"}
+                title={t("webSearch")}
                 path="/v1/search"
                 models={searchProviders.map((p) => ({ id: p.id, owned_by: p.id, type: "search" }))}
                 copy={copy}
@@ -2008,7 +2011,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
           <div className="flex items-center gap-2 mb-3">
             <span className="material-symbols-outlined text-sm text-amber-400">build</span>
             <h3 className="text-xs font-semibold text-text-muted uppercase tracking-wider">
-              {t("categoryUtility") || "Utility & Management"}
+              {t("categoryUtility")}
             </h3>
             <div className="flex-1 h-px bg-border/50" />
           </div>
@@ -2041,7 +2044,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="view_list"
               iconColor="text-teal-500"
               iconBg="bg-teal-500/10"
-              title={t("batchApi") || "Batch API"}
+              title={t("batchApi")}
               path="/v1/batches"
               models={null}
               badge="OpenAI"
@@ -2053,7 +2056,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="folder"
               iconColor="text-yellow-500"
               iconBg="bg-yellow-500/10"
-              title={t("filesApi") || "Files API"}
+              title={t("filesApi")}
               path="/v1/files"
               models={null}
               copy={copy}
@@ -2064,7 +2067,7 @@ export default function APIPageClient({ machineId }: Readonly<APIPageClientProps
               icon="list"
               iconColor="text-teal-500"
               iconBg="bg-teal-500/10"
-              title={t("listModels") || "List Models"}
+              title={t("listModels")}
               path="/v1/models"
               models={null}
               copy={copy}

src/app/(dashboard)/dashboard/endpoint/components/A2ADashboard.tsx

@@ -429,7 +429,7 @@ export default function A2ADashboardPage() {
                   <th className="text-left py-2 pr-2">{t("tableTask")}</th>
                   <th className="text-left py-2 pr-2">{t("tableSkill")}</th>
                   <th className="text-left py-2 pr-2">{t("tableState")}</th>
-                  <th className="text-left py-2 pr-2">{t("tablePhase") || "FSM Status"}</th>
+                  <th className="text-left py-2 pr-2">{t("tablePhase")}</th>
                   <th className="text-left py-2 pr-2">{t("tableUpdated")}</th>
                   <th className="text-left py-2">{t("tableActions")}</th>
                 </tr>

src/app/(dashboard)/dashboard/endpoint/components/NotionSourceCard.tsx

@@ -12,6 +12,18 @@ export default function NotionSourceCard() {
   const [message, setMessage] = useState<{ type: "success" | "error"; text: string } | null>(null);
   const [expanded, setExpanded] = useState(false);
 
+  const translateOrFallback = (key: string, fallback: string) => {
+    try {
+      const translated = t(key as never);
+      if (!translated || translated === key || translated === `endpoint.${key}`) {
+        return fallback;
+      }
+      return translated;
+    } catch {
+      return fallback;
+    }
+  };
+
   const fetchConfig = useCallback(async () => {
     try {
       const res = await fetch("/api/settings/notion");
@@ -37,7 +49,10 @@ export default function NotionSourceCard() {
 
   const handleSaveToken = async () => {
     if (!token.trim()) {
-      setMessage({ type: "error", text: "Please enter a Notion integration token" });
+      setMessage({
+        type: "error",
+        text: translateOrFallback("notionEnterToken", "Please enter a Notion integration token"),
+      });
       return;
     }
     setBusy(true);
@@ -53,11 +68,20 @@ export default function NotionSourceCard() {
         setConnected(true);
         setMessage({ type: "success", text: data.message });
       } else {
-        setMessage({ type: "error", text: data.error ?? "Failed to connect" });
+        setMessage({
+          type: "error",
+          text: data.error ?? translateOrFallback("notionConnectFailed", "Failed to connect"),
+        });
         setConnected(false);
       }
     } catch (err) {
-      setMessage({ type: "error", text: err instanceof Error ? err.message : "Connection failed" });
+      setMessage({
+        type: "error",
+        text:
+          err instanceof Error
+            ? err.message
+            : translateOrFallback("notionConnectionFailed", "Connection failed"),
+      });
     } finally {
       setBusy(false);
     }
@@ -74,10 +98,19 @@ export default function NotionSourceCard() {
         setToken("");
         setMessage({ type: "success", text: data.message });
       } else {
-        setMessage({ type: "error", text: data.error ?? "Failed to disconnect" });
+        setMessage({
+          type: "error",
+          text: data.error ?? translateOrFallback("notionDisconnectFailed", "Failed to disconnect"),
+        });
       }
     } catch (err) {
-      setMessage({ type: "error", text: err instanceof Error ? err.message : "Disconnect failed" });
+      setMessage({
+        type: "error",
+        text:
+          err instanceof Error
+            ? err.message
+            : translateOrFallback("notionDisconnectFailed", "Disconnect failed"),
+      });
     } finally {
       setBusy(false);
     }
@@ -97,11 +130,16 @@ export default function NotionSourceCard() {
             <div className="flex items-center gap-2 flex-wrap">
               <span className="font-semibold text-sm">Notion</span>
               <Badge variant={connected ? "success" : "default"}>
-                {connected ? "Connected" : "Not connected"}
+                {connected
+                  ? translateOrFallback("notionConnected", "Connected")
+                  : translateOrFallback("notionNotConnected", "Not connected")}
               </Badge>
             </div>
             <p className="text-xs text-text-muted mt-0.5">
-              Search, read, query, and write to Notion through routed AI models
+              {translateOrFallback(
+                "notionDescription",
+                "Search, read, query, and write to Notion through routed AI models"
+              )}
             </p>
           </div>
           <span
@@ -131,7 +169,10 @@ export default function NotionSourceCard() {
             {!connected ? (
               <div className="flex flex-col gap-2">
                 <label className="text-xs text-text-muted font-medium">
-                  Notion Internal Integration Token
+                  {translateOrFallback(
+                    "notionIntegrationToken",
+                    "Notion Internal Integration Token"
+                  )}
                 </label>
                 <div className="flex gap-2">
                   <Input
@@ -143,11 +184,14 @@ export default function NotionSourceCard() {
                     className="font-mono text-sm flex-1"
                   />
                   <Button onClick={handleSaveToken} loading={busy} variant="primary" size="sm">
-                    Connect
+                    {translateOrFallback("notionConnect", "Connect")}
                   </Button>
                 </div>
                 <p className="text-[10px] text-text-muted">
-                  Create an Internal Integration at{" "}
+                  {translateOrFallback(
+                    "notionIntegrationHelp",
+                    "Create an Internal Integration at"
+                  )}{" "}
                   <code className="text-primary font-mono bg-surface/80 px-1 rounded">
                     https://www.notion.so/profile/integrations
                   </code>
@@ -156,7 +200,10 @@ export default function NotionSourceCard() {
             ) : (
               <div className="flex items-center gap-2">
                 <span className="text-xs text-text-muted flex-1">
-                  Token configured. Notion tools are available via MCP.
+                  {translateOrFallback(
+                    "notionTokenConfigured",
+                    "Token configured. Notion tools are available via MCP."
+                  )}
                 </span>
                 <Button
                   onClick={handleDisconnect}
@@ -165,7 +212,7 @@ export default function NotionSourceCard() {
                   size="sm"
                   className="border-red-500/30! text-red-400! hover:bg-red-500/10!"
                 >
-                  Disconnect
+                  {translateOrFallback("notionDisconnect", "Disconnect")}
                 </Button>
               </div>
             )}

src/app/(dashboard)/dashboard/plugins/page.tsx

@@ -1,7 +1,7 @@
 "use client";
 
 import { useState, useEffect, useCallback } from "react";
-import { Card, Button, EmptyState, Badge } from "@/shared/components";
+import { Card, Button, EmptyState } from "@/shared/components";
 import { useNotificationStore } from "@/store/notificationStore";
 import { useTranslations } from "next-intl";
 
@@ -19,10 +19,6 @@ export default function PluginsPage() {
   const { addNotification } = useNotificationStore();
   const t = useTranslations("plugins");
   const [plugins, setPlugins] = useState<PluginInfo[]>([]);
-  const [activeTab, setActiveTab] = useState<"installed" | "marketplace">("installed");
-  const [marketplacePlugins, setMarketplacePlugins] = useState<any[]>([]);
-  const [marketplaceUrl, setMarketplaceUrl] = useState("");
-  const [savingUrl, setSavingUrl] = useState(false);
   const [loading, setLoading] = useState(true);
   const [scanning, setScanning] = useState(false);
 
@@ -42,51 +38,7 @@ export default function PluginsPage() {
 
   useEffect(() => {
     fetchPlugins();
-    fetch("/api/settings")
-      .then((res) => res.ok ? res.json() : null)
-      .then((data) => {
-        if (data?.pluginMarketplaceUrl) setMarketplaceUrl(data.pluginMarketplaceUrl);
-      })
-      .catch(() => {});
   }, [fetchPlugins]);
-  
-  const fetchMarketplace = useCallback(async () => {
-    try {
-      const res = await fetch("/api/plugins/marketplace");
-      if (res.ok) {
-        const data = await res.json();
-        setMarketplacePlugins(data.plugins || []);
-      }
-    } catch {}
-  }, []);
-  
-  useEffect(() => {
-    if (activeTab === "marketplace") {
-      fetchMarketplace();
-    }
-  }, [activeTab, fetchMarketplace]);
-  
-  const handleSaveUrl = async () => {
-    setSavingUrl(true);
-    try {
-      const res = await fetch("/api/settings", {
-        method: "PATCH",
-        headers: { "Content-Type": "application/json" },
-        body: JSON.stringify({ pluginMarketplaceUrl: marketplaceUrl || null }),
-      });
-      if (!res.ok) {
-        const errData = await res.json().catch(() => null);
-        addNotification({ type: "error", message: errData?.error || "Failed to save" });
-        return;
-      }
-      addNotification({ type: "success", message: t("marketplaceUrlSaved") });
-      await fetchMarketplace();
-    } catch {
-      addNotification({ type: "error", message: t("saveConfigurationFailed") });
-    } finally {
-      setSavingUrl(false);
-    }
-  };
 
   const handleScan = async () => {
     setScanning(true);
@@ -137,127 +89,58 @@ export default function PluginsPage() {
     <div className="space-y-6 p-6">
       <div className="flex items-center justify-between">
         <h1 className="text-2xl font-bold">{t("title")}</h1>
-        <div className="flex gap-2">
-          <Button variant={activeTab === "installed" ? "primary" : "secondary"} onClick={() => setActiveTab("installed")}>
-            {t("installedTab")}
-          </Button>
-          <Button variant={activeTab === "marketplace" ? "primary" : "secondary"} onClick={() => setActiveTab("marketplace")}>
-            {t("marketplaceTab")}
-          </Button>
-        </div>
       </div>
 
-      {activeTab === "marketplace" && (
-        <Card className="p-4 flex gap-4 items-end bg-gray-50">
-          <div className="flex-1">
-            <label className="block text-sm font-medium text-gray-700 mb-1">{t("marketplaceUrlLabel")}</label>
-            <input
-              type="text"
-              className="w-full rounded border-gray-300 p-2"
-              placeholder={t("marketplaceUrlPlaceholder")}
-              value={marketplaceUrl}
-              onChange={(e) => setMarketplaceUrl(e.target.value)}
-            />
-          </div>
-          <Button onClick={handleSaveUrl} disabled={savingUrl}>
-            {t("saveMarketplaceUrl")}
-          </Button>
-        </Card>
-      )}
-
-      {activeTab === "installed" ? (
-        <>
-          <div className="flex items-center justify-end">
-            <Button onClick={handleScan} disabled={scanning}>
-              {scanning ? t("scanning") : t("scanForPlugins")}
-            </Button>
-          </div>
-          {plugins.length === 0 ? (
-            <EmptyState
-              title={t("noPlugins")}
-              description={t("noPluginsDescription")}
-            />
-          ) : (
-            <div className="grid gap-4">
-              {plugins.map((plugin) => (
-                <Card key={plugin.name} className="p-4">
-                  <div className="flex items-center justify-between">
-                    <div>
-                      <h3 className="font-semibold">{plugin.name}</h3>
-                      <p className="text-sm text-gray-500">
-                        v{plugin.version}
-                        {plugin.author ? ` by ${plugin.author}` : ""}
-                        {plugin.description ? ` — ${plugin.description}` : ""}
-                      </p>
-                      <div className="mt-1 flex gap-1">
-                        {plugin.hooks.map((hook) => (
-                          <span
-                            key={hook}
-                            className="rounded bg-blue-100 px-2 py-0.5 text-xs text-blue-700"
-                          >
-                            {hook}
-                          </span>
-                        ))}
-                      </div>
-                    </div>
-                    <div className="flex gap-2">
-                      <Button
-                        variant={plugin.enabled ? "secondary" : "primary"}
-                        onClick={() => handleToggle(plugin.name, !plugin.enabled)}
-                      >
-                        {plugin.enabled ? t("deactivate") : t("activate")}
-                      </Button>
-                      <Button
-                        variant="danger"
-                        onClick={() => handleUninstall(plugin.name)}
-                      >
-                        {t("uninstall")}
-                      </Button>
-                    </div>
-                  </div>
-                </Card>
-              ))}
-            </div>
-          )}
-        </>
+      <div className="flex items-center justify-end">
+        <Button onClick={handleScan} disabled={scanning}>
+          {scanning ? t("scanning") : t("scanForPlugins")}
+        </Button>
+      </div>
+      {plugins.length === 0 ? (
+        <EmptyState
+          title={t("noPlugins")}
+          description={t("noPluginsDescription")}
+        />
       ) : (
         <div className="grid gap-4">
-          {marketplacePlugins.length === 0 ? (
-             <div className="text-gray-500 py-4">{t("marketplaceEmpty")}</div>
-          ) : (
-            marketplacePlugins.map((plugin) => (
-              <Card key={plugin.name} className="p-4">
-                <div className="flex items-center justify-between">
-                  <div>
-                    <h3 className="font-semibold flex items-center gap-2">
-                      {plugin.name}
-                      {plugin.verified && <Badge variant="success">{t("verified")}</Badge>}
-                    </h3>
-                    <p className="text-sm text-gray-500">
-                      v{plugin.version} by {plugin.author} — {plugin.description}
-                    </p>
-                    <div className="mt-1 flex gap-1">
-                      {plugin.tags?.map((tag: string) => (
-                        <span key={tag} className="rounded bg-gray-100 px-2 py-0.5 text-xs text-gray-600">
-                          {tag}
-                        </span>
-                      ))}
-                    </div>
-                  </div>
-                  <div className="flex gap-2">
-                    <Button
-                      variant="primary"
-                      onClick={() => {
-                        addNotification({ type: "info", message: t("marketplaceInstallComingSoon") });
-                      }}
-                    >
-                      {t("install")}
-                    </Button>
+          {plugins.map((plugin) => (
+            <Card key={plugin.name} className="p-4">
+              <div className="flex items-center justify-between">
+                <div>
+                  <h3 className="font-semibold">{plugin.name}</h3>
+                  <p className="text-sm text-gray-500">
+                    v{plugin.version}
+                    {plugin.author ? ` by ${plugin.author}` : ""}
+                    {plugin.description ? ` — ${plugin.description}` : ""}
+                  </p>
+                  <div className="mt-1 flex gap-1">
+                    {plugin.hooks.map((hook) => (
+                      <span
+                        key={hook}
+                        className="rounded bg-blue-100 px-2 py-0.5 text-xs text-blue-700"
+                      >
+                        {hook}
+                      </span>
+                    ))}
                   </div>
                 </div>
-              </Card>
-            ))
-          )}
+                <div className="flex gap-2">
+                  <Button
+                    variant={plugin.enabled ? "secondary" : "primary"}
+                    onClick={() => handleToggle(plugin.name, !plugin.enabled)}
+                  >
+                    {plugin.enabled ? t("deactivate") : t("activate")}
+                  </Button>
+                  <Button
+                    variant="danger"
+                    onClick={() => handleUninstall(plugin.name)}
+                  >
+                    {t("uninstall")}
+                  </Button>
+                </div>
+              </div>
+            </Card>
+          ))}
         </div>
       )}
     </div>

src/app/(dashboard)/dashboard/providers/[id]/components/ConnectionRow.tsx

@@ -15,11 +15,7 @@ import {
   getCodexEffectiveServiceTier,
   type CodexGlobalServiceMode,
 } from "@/lib/providers/codexFastTier";
-import {
-  normalizeCodexLimitPolicy,
-  providerText,
-  ERROR_TYPE_LABELS,
-} from "../providerPageHelpers";
+import { normalizeCodexLimitPolicy, providerText, ERROR_TYPE_LABELS } from "../providerPageHelpers";
 import { getCodexPlanLabel } from "../codexPlanLabel";
 import ProviderQuotaVisibilityToggle from "./ProviderQuotaVisibilityToggle";
 
@@ -245,7 +241,7 @@ function getStatusPresentation(
   if (errorType === "account_deactivated") {
     return {
       statusVariant: "error",
-      statusLabel: t("statusDeactivated", "Deactivated"),
+      statusLabel: providerText(t, "statusDeactivated", "Deactivated"),
       errorType,
       errorBadge,
       errorTextClass: "text-red-600 font-bold",
@@ -300,7 +296,7 @@ function getStatusPresentation(
   if (errorType === "banned") {
     return {
       statusVariant: "error",
-      statusLabel: t("statusBanned", "Banned (403)"),
+      statusLabel: providerText(t, "statusBanned", "Banned (403)"),
       errorType,
       errorBadge,
       errorTextClass: "text-red-600 font-bold",
@@ -310,7 +306,7 @@ function getStatusPresentation(
   if (errorType === "credits_exhausted") {
     return {
       statusVariant: "warning",
-      statusLabel: t("statusCreditsExhausted", "Out of Credits"),
+      statusLabel: providerText(t, "statusCreditsExhausted", "Out of Credits"),
       errorType,
       errorBadge,
       errorTextClass: "text-amber-500",
@@ -391,22 +387,10 @@ export default function ConnectionRow({
         t("oauthAccount")
       )
     : connection.name;
-  const applyCodexAuthLabel =
-    typeof t.has === "function" && t.has("applyCodexAuthLocal")
-      ? t("applyCodexAuthLocal")
-      : "Apply auth";
-  const exportCodexAuthLabel =
-    typeof t.has === "function" && t.has("exportCodexAuthFile")
-      ? t("exportCodexAuthFile")
-      : "Export auth";
-  const applyClaudeAuthLabel =
-    typeof t.has === "function" && t.has("applyClaudeAuthLocal")
-      ? t("applyClaudeAuthLocal")
-      : "Apply auth";
-  const exportClaudeAuthLabel =
-    typeof t.has === "function" && t.has("exportClaudeAuthFile")
-      ? t("exportClaudeAuthFile")
-      : "Export auth";
+  const applyCodexAuthLabel = providerText(t, "applyCodexAuthLocal", "Apply auth");
+  const exportCodexAuthLabel = providerText(t, "exportCodexAuthFile", "Export auth");
+  const applyClaudeAuthLabel = providerText(t, "applyClaudeAuthLocal", "Apply auth");
+  const exportClaudeAuthLabel = providerText(t, "exportClaudeAuthFile", "Export auth");
   // Use useState + useEffect for impure Date.now() to avoid calling during render
   const [isCooldown, setIsCooldown] = useState(false);
   // T12: token expiry status — lazy init avoids calling Date.now() during render;

src/app/(dashboard)/dashboard/providers/[id]/components/CoolingConnectionsPanel.tsx

@@ -22,8 +22,10 @@
  */
 
 import { useEffect, useState } from "react";
+import { useTranslations } from "next-intl";
 import { formatResetCountdown } from "@/shared/utils/formatting";
 import type { ConnectionRowConnection } from "./ConnectionRow";
+import { providerText } from "../providerPageHelpers";
 
 export interface CoolingConnectionsPanelProps {
   readonly connections: readonly ConnectionRowConnection[];
@@ -37,6 +39,7 @@ function isCoolingNow(connection: ConnectionRowConnection, now: number): boolean
 
 export default function CoolingConnectionsPanel(props: CoolingConnectionsPanelProps) {
   const { connections } = props;
+  const t = useTranslations("providers");
   // Tick once per second so the human-readable countdown updates.
   const [now, setNow] = useState<number>(() => Date.now());
   useEffect(() => {
@@ -58,12 +61,17 @@ export default function CoolingConnectionsPanel(props: CoolingConnectionsPanelPr
           className="inline-block h-2 w-2 animate-pulse rounded-full bg-amber-500"
         />
         <h3 className="text-sm font-medium text-amber-700 dark:text-amber-300">
-          Currently cooling ({cooling.length})
+          {providerText(t, "coolingConnectionsTitle", "Currently cooling ({count})", {
+            count: cooling.length,
+          })}
         </h3>
       </div>
       <p className="mb-3 text-xs text-muted-foreground">
-        These connections returned a 429 (rate-limit) on their last request. OmniRoute will skip
-        them until the timer expires — no manual disable required.
+        {providerText(
+          t,
+          "coolingConnectionsDescription",
+          "These connections returned a 429 (rate-limit) on their last request. OmniRoute will skip them until the timer expires — no manual disable required."
+        )}
       </p>
       <ul className="space-y-1">
         {cooling.map((c) => {
@@ -72,7 +80,9 @@ export default function CoolingConnectionsPanel(props: CoolingConnectionsPanelPr
             c.displayName ||
             c.name ||
             c.email ||
-            (c.id ? `connection ${c.id.slice(0, 8)}` : "connection");
+            (c.id
+              ? `${providerText(t, "connectionFallback", "connection")} ${c.id.slice(0, 8)}`
+              : providerText(t, "connectionFallback", "connection"));
           return (
             <li
               key={c.id ?? label}

src/app/(dashboard)/dashboard/providers/[id]/components/CustomModelsSection.tsx

@@ -21,6 +21,7 @@ import {
   effectivePreserveForProtocol,
   effectiveUpstreamHeadersForProtocol,
   formatProviderModelsErrorResponse,
+  providerText,
   targetFormatBadgeI18nKey,
   type CompatModelRow,
   type CompatByProtocolMap,
@@ -285,17 +286,32 @@ export default function CustomModelsSection({
 
       if (!res.ok) {
         const detail = await formatProviderModelsErrorResponse(res);
-        throw new Error(detail || "Failed to save model endpoint settings");
+        throw new Error(
+          detail ||
+            providerText(
+              t,
+              "failedSaveModelEndpointSettings",
+              "Failed to save model endpoint settings"
+            )
+        );
       }
 
       await fetchCustomModels();
       onModelsChanged?.();
-      notify.success("Saved model endpoint settings");
+      notify.success(
+        providerText(t, "savedModelEndpointSettings", "Saved model endpoint settings")
+      );
       cancelEdit();
     } catch (e) {
       console.error("Failed to save custom model:", e);
       notify.error(
-        e instanceof Error && e.message ? e.message : "Failed to save model endpoint settings"
+        e instanceof Error && e.message
+          ? e.message
+          : providerText(
+              t,
+              "failedSaveModelEndpointSettings",
+              "Failed to save model endpoint settings"
+            )
       );
     } finally {
       setSavingModelId(null);
@@ -305,7 +321,9 @@ export default function CustomModelsSection({
   const saveEdit = async (modelId: string) => {
     if (!editingModelId || editingModelId !== modelId) return;
     if (!editingEndpoints.length) {
-      notify.error("Select at least one supported endpoint");
+      notify.error(
+        providerText(t, "selectSupportedEndpoint", "Select at least one supported endpoint")
+      );
       return;
     }
 
@@ -429,7 +447,7 @@ export default function CustomModelsSection({
                     : ep === "embeddings"
                       ? `📐 ${t("supportedEndpointEmbeddings")}`
                       : ep === "rerank"
-                        ? "Rerank"
+                        ? providerText(t, "rerankEndpoint", "Rerank")
                         : ep === "images"
                           ? `🖼️ ${t("supportedEndpointImages")}`
                           : `🔊 ${t("supportedEndpointAudio")}`}
@@ -661,7 +679,7 @@ export default function CustomModelsSection({
                                   : ep === "embeddings"
                                     ? `📐 ${t("supportedEndpointEmbeddings")}`
                                     : ep === "rerank"
-                                      ? "Rerank"
+                                      ? providerText(t, "rerankEndpoint", "Rerank")
                                       : ep === "images"
                                         ? `🖼️ ${t("supportedEndpointImages")}`
                                         : `🔊 ${t("supportedEndpointAudio")}`}

src/app/(dashboard)/dashboard/providers/[id]/components/KimiCodeAuthMethodModal.tsx

@@ -1,7 +1,7 @@
 "use client";
 
 import { Modal } from "@/shared/components";
-import type { ProviderMessageTranslator } from "../providerPageHelpers";
+import { providerText, type ProviderMessageTranslator } from "../providerPageHelpers";
 
 type KimiCodeAuthMethodModalProps = {
   isOpen: boolean;
@@ -50,7 +50,9 @@ export default function KimiCodeAuthMethodModal({
           <div className="flex items-start gap-3">
             <span className="material-symbols-outlined mt-0.5 text-primary">key</span>
             <div className="min-w-0 flex-1">
-              <h3 className="mb-1 font-semibold">Kimi Code API Key</h3>
+              <h3 className="mb-1 font-semibold">
+                {providerText(t, "kimiCodeApiKeyLabel", "Kimi Code API Key")}
+              </h3>
               <p className="text-sm text-text-muted">{t("apiKeySecure")}</p>
             </div>
           </div>

src/app/(dashboard)/dashboard/providers/[id]/components/ModelCompatPopover.tsx

@@ -13,6 +13,7 @@ import {
   UPSTREAM_HEADERS_UI_MAX,
   headerRowsToRecord,
   compatProtocolLabelKey,
+  providerText,
   type HeaderDraftRow,
 } from "../providerPageHelpers";
 
@@ -353,7 +354,7 @@ export default function ModelCompatPopover({
               {/* Param filters — model-level block/allow (#6625) */}
               <div className="mt-4 space-y-2.5">
                 <label className="block text-[11px] font-semibold text-text-main">
-                  {t("compatParamFiltersLabel") ?? "Param Filters"}
+                  {providerText(t, "compatParamFiltersLabel", "Param Filters")}
                 </label>
                 <div>
                   <input
@@ -369,7 +370,11 @@ export default function ModelCompatPopover({
                     className="mb-1 w-full rounded-lg border border-zinc-200 bg-white px-2.5 py-1.5 text-[11px] font-mono text-text-main placeholder:text-text-muted focus:border-primary focus:outline-none focus:ring-1 focus:ring-primary/30 dark:border-zinc-600 dark:bg-zinc-900"
                   />
                   <p className="text-[10px] text-text-muted">
-                    {t("compatBlockedParamsHint") ?? "Blocked params (stripped from requests)"}
+                    {providerText(
+                      t,
+                      "compatBlockedParamsHint",
+                      "Blocked params (stripped from requests)"
+                    )}
                     {paramSaving && ` ● ${t("compatSaving")}`}
                   </p>
                 </div>
@@ -387,7 +392,11 @@ export default function ModelCompatPopover({
                     className="mb-1 w-full rounded-lg border border-zinc-200 bg-white px-2.5 py-1.5 text-[11px] font-mono text-text-main placeholder:text-text-muted focus:border-primary focus:outline-none focus:ring-1 focus:ring-primary/30 dark:border-zinc-600 dark:bg-zinc-900"
                   />
                   <p className="text-[10px] text-text-muted">
-                    {t("compatAllowedParamsHint") ?? "Allowed params (re-added after deny)"}
+                    {providerText(
+                      t,
+                      "compatAllowedParamsHint",
+                      "Allowed params (re-added after deny)"
+                    )}
                   </p>
                 </div>
               </div>

src/app/(dashboard)/dashboard/providers/[id]/components/ModelRow.tsx

@@ -411,7 +411,7 @@ export default function ModelRow({
                 : testStatus === "ok"
                   ? "OK"
                   : testStatus === "error"
-                    ? "Error"
+                    ? providerText(t, "errorShort", "Error")
                     : t("testModel")
             }
           >

src/app/(dashboard)/dashboard/providers/[id]/components/NoAuthProviderControls.tsx

@@ -5,6 +5,7 @@ import { useTranslations } from "next-intl";
 import { NoAuthAccountCard, NoAuthProviderCard } from "@/shared/components";
 import { getProviderAlias, supportsNoAuthProviderProxy } from "@/shared/constants/providers";
 import { useNotificationStore } from "@/store/notificationStore";
+import { providerText } from "../providerPageHelpers";
 
 const ACCOUNT_PROVIDER_NAMES: Record<string, string> = {
   mimocode: "MiMoCode",
@@ -123,7 +124,10 @@ export default function NoAuthProviderControls({
                 const res = await fetch("/api/dahl/tokens", { method: "POST" });
                 const data = await res.json();
                 if (!res.ok || !data.token) {
-                  throw new Error(data?.error || "Failed to create Dahl token");
+                  throw new Error(
+                    data?.error ||
+                      providerText(t, "createDahlTokenFailed", "Failed to create Dahl token")
+                  );
                 }
                 return data.token as string;
               }

src/app/(dashboard)/dashboard/providers/[id]/components/PassthroughModelRow.tsx

@@ -190,7 +190,7 @@ export default function PassthroughModelRow({
                   : testStatus === "ok"
                     ? "OK"
                     : testStatus === "error"
-                      ? "Error"
+                      ? providerText(t, "errorShort", "Error")
                       : t("testModel")
               }
             >

src/app/(dashboard)/dashboard/providers/[id]/components/ProviderPlaygroundPanel.tsx

@@ -6,6 +6,7 @@
 // a single-kind panel or the LlmChatCard for standard LLM providers.
 
 import { useState } from "react";
+import { useTranslations } from "next-intl";
 import { LlmChatCard } from "@/app/(dashboard)/dashboard/media-providers/components/LlmChatCard";
 import { ServiceKindTabs } from "@/app/(dashboard)/dashboard/media-providers/components/ServiceKindTabs";
 import { EmbeddingExampleCard } from "@/app/(dashboard)/dashboard/media-providers/components/EmbeddingExampleCard";
@@ -18,6 +19,7 @@ import { VideoExampleCard } from "@/app/(dashboard)/dashboard/media-providers/co
 import { MusicExampleCard } from "@/app/(dashboard)/dashboard/media-providers/components/MusicExampleCard";
 import type { ServiceKind } from "@/shared/constants/providers";
 import { AI_PROVIDERS } from "@/shared/constants/providers";
+import { providerText } from "../providerPageHelpers";
 
 export const MEDIA_SERVICE_KINDS: ServiceKind[] = [
   "embedding",
@@ -56,12 +58,12 @@ export function renderKindPanel(kind: ServiceKind, providerId: string): JSX.Elem
 }
 
 export default function ProviderPlaygroundPanel({ providerId }: { providerId: string }) {
+  const t = useTranslations("providers");
   // Resolve serviceKinds from AI_PROVIDERS.
   // For providers without explicit serviceKinds (most LLM providers), we infer
   // "llm" as the default.
   const providerEntry = AI_PROVIDERS[providerId as keyof typeof AI_PROVIDERS] as
-    | (Record<string, unknown> & { serviceKinds?: string[] })
-    | undefined;
+    (Record<string, unknown> & { serviceKinds?: string[] }) | undefined;
 
   const rawKinds: string[] = providerEntry?.serviceKinds ?? [];
 
@@ -93,7 +95,7 @@ export default function ProviderPlaygroundPanel({ providerId }: { providerId: st
 
   return (
     <div className="flex flex-col gap-3">
-      <h2 className="text-lg font-semibold">Playground</h2>
+      <h2 className="text-lg font-semibold">{providerText(t, "playgroundTitle", "Playground")}</h2>
       <ServiceKindTabs
         kinds={playgroundableKinds}
         activeKind={activeKind}

src/app/(dashboard)/dashboard/providers/[id]/components/modals/EditCompatibleNodeModal.tsx

@@ -3,6 +3,7 @@ import { useState, useEffect } from "react";
 import { useTranslations } from "next-intl";
 import { Button, Badge, Input, Modal, Select } from "@/shared/components";
 import { CC_COMPATIBLE_DEFAULT_CHAT_PATH } from "../../providerDetailConstants";
+import { providerText } from "../../providerPageHelpers";
 interface EditCompatibleNodeModalNode {
   id?: string;
   name?: string;
@@ -45,9 +46,11 @@ export default function EditCompatibleNodeModal({
   const [checkKey, setCheckKey] = useState("");
   const [checkModelId, setCheckModelId] = useState("");
   const [validating, setValidating] = useState(false);
-  const [validationResult, setValidationResult] = useState<
-    null | { valid: boolean; error?: string | null; method?: string | null }
-  >(null);
+  const [validationResult, setValidationResult] = useState<null | {
+    valid: boolean;
+    error?: string | null;
+    method?: string | null;
+  }>(null);
   const [showAdvanced, setShowAdvanced] = useState(false);
 
   useEffect(() => {
@@ -130,7 +133,10 @@ export default function EditCompatibleNodeModal({
         method: data.method ?? null,
       });
     } catch {
-      setValidationResult({ valid: false, error: "Network error" });
+      setValidationResult({
+        valid: false,
+        error: providerText(t, "networkError", "Network error"),
+      });
     } finally {
       setValidating(false);
     }

src/app/(dashboard)/dashboard/providers/[id]/hooks/useCommandCodeAuth.ts

@@ -1,5 +1,6 @@
 import { useState, useEffect, useCallback, useRef } from "react";
-import { type CommandCodeAuthFlowState } from "../providerPageHelpers";
+import { useTranslations } from "next-intl";
+import { providerText, type CommandCodeAuthFlowState } from "../providerPageHelpers";
 
 export type UseCommandCodeAuthParams = {
   providerId: string;
@@ -15,6 +16,7 @@ export function useCommandCodeAuth({
   setShowAddApiKeyModal,
   notify,
 }: UseCommandCodeAuthParams) {
+  const t = useTranslations("providers");
   const [commandCodeAuthState, setCommandCodeAuthState] = useState<CommandCodeAuthFlowState>({
     phase: "idle",
     state: "",
@@ -62,7 +64,7 @@ export function useCommandCodeAuth({
       setCommandCodeAuthState((current) => ({
         ...current,
         phase: "applying",
-        message: "Applying browser-approved key…",
+        message: providerText(t, "commandCodeApplyingKey", "Applying browser-approved key…"),
       }));
 
       try {
@@ -74,7 +76,9 @@ export function useCommandCodeAuth({
         const data = await res.json().catch(() => ({}));
 
         if (!res.ok) {
-          const errorMessage = data.error || "Failed to apply Command Code auth";
+          const errorMessage =
+            data.error ||
+            providerText(t, "commandCodeApplyFailed", "Failed to apply Command Code auth");
           setCommandCodeAuthState((current) => ({
             ...current,
             phase: "error",
@@ -87,26 +91,30 @@ export function useCommandCodeAuth({
         setCommandCodeAuthState((current) => ({
           ...current,
           phase: "applied",
-          message: "Command Code connected",
+          message: providerText(t, "commandCodeConnected", "Command Code connected"),
         }));
         commandCodeAuthWindowRef.current?.close?.();
         commandCodeAuthWindowRef.current = null;
         await fetchConnections();
         handleCloseAddApiKeyModal();
-        notify.success("Command Code connection added");
+        notify.success(
+          providerText(t, "commandCodeConnectionAdded", "Command Code connection added")
+        );
         return true;
       } catch (error) {
         console.error("Error applying Command Code auth:", error);
         setCommandCodeAuthState((current) => ({
           ...current,
           phase: "error",
-          message: "Failed to apply Command Code auth",
+          message: providerText(t, "commandCodeApplyFailed", "Failed to apply Command Code auth"),
         }));
-        notify.error("Failed to apply Command Code auth");
+        notify.error(
+          providerText(t, "commandCodeApplyFailed", "Failed to apply Command Code auth")
+        );
         return false;
       }
     },
-    [fetchConnections, handleCloseAddApiKeyModal, notify]
+    [fetchConnections, handleCloseAddApiKeyModal, notify, t]
   );
 
   const handleStartCommandCodeAuth = useCallback(async () => {
@@ -124,7 +132,7 @@ export function useCommandCodeAuth({
       authUrl: "",
       callbackUrl: "",
       expiresAt: null,
-      message: "Opening Command Code Studio…",
+      message: providerText(t, "commandCodeOpeningStudio", "Opening Command Code Studio…"),
     });
 
     try {
@@ -135,7 +143,9 @@ export function useCommandCodeAuth({
       const data = await res.json().catch(() => ({}));
 
       if (!res.ok || !data.state || !data.authUrl) {
-        const errorMessage = data.error || "Failed to start Command Code auth";
+        const errorMessage =
+          data.error ||
+          providerText(t, "commandCodeStartFailed", "Failed to start Command Code auth");
         setCommandCodeAuthState((current) => ({
           ...current,
           phase: "error",
@@ -152,7 +162,11 @@ export function useCommandCodeAuth({
         authUrl: data.authUrl,
         callbackUrl: data.callbackUrl || "",
         expiresAt: data.expiresAt || null,
-        message: "Open the auth URL, approve access, then paste the returned key/JSON/URL below…",
+        message: providerText(
+          t,
+          "commandCodeApprovalInstructions",
+          "Open the auth URL, approve access, then paste the returned key/JSON/URL below…"
+        ),
       });
 
       if (popup) {
@@ -169,9 +183,19 @@ export function useCommandCodeAuth({
           setCommandCodeAuthState((current) => ({
             ...current,
             phase: "error",
-            message: "Popup blocked. Please allow popups and try Command Code Connect again.",
+            message: providerText(
+              t,
+              "commandCodePopupBlocked",
+              "Popup blocked. Please allow popups and try Command Code Connect again."
+            ),
           }));
-          notify.error("Popup blocked. Please allow popups and try Command Code Connect again.");
+          notify.error(
+            providerText(
+              t,
+              "commandCodePopupBlocked",
+              "Popup blocked. Please allow popups and try Command Code Connect again."
+            )
+          );
           return;
         }
         commandCodeAuthWindowRef.current = fallbackPopup;
@@ -183,11 +207,11 @@ export function useCommandCodeAuth({
           setCommandCodeAuthState((current) => ({
             ...current,
             phase: "expired",
-            message: "Command Code link expired",
+            message: providerText(t, "commandCodeLinkExpired", "Command Code link expired"),
           }));
           commandCodeAuthWindowRef.current?.close?.();
           commandCodeAuthWindowRef.current = null;
-          notify.error("Command Code auth expired");
+          notify.error(providerText(t, "commandCodeAuthExpired", "Command Code auth expired"));
           clearCommandCodeAuthTimer();
           return;
         }
@@ -206,11 +230,11 @@ export function useCommandCodeAuth({
             setCommandCodeAuthState((current) => ({
               ...current,
               phase: "expired",
-              message: "Command Code link expired",
+              message: providerText(t, "commandCodeLinkExpired", "Command Code link expired"),
             }));
             commandCodeAuthWindowRef.current?.close?.();
             commandCodeAuthWindowRef.current = null;
-            notify.error("Command Code auth expired");
+            notify.error(providerText(t, "commandCodeAuthExpired", "Command Code auth expired"));
             clearCommandCodeAuthTimer();
             return;
           }
@@ -219,13 +243,15 @@ export function useCommandCodeAuth({
             setCommandCodeAuthState((current) => ({
               ...current,
               phase: "applied",
-              message: "Command Code connected",
+              message: providerText(t, "commandCodeConnected", "Command Code connected"),
             }));
             commandCodeAuthWindowRef.current?.close?.();
             commandCodeAuthWindowRef.current = null;
             await fetchConnections();
             handleCloseAddApiKeyModal();
-            notify.success("Command Code connection added");
+            notify.success(
+              providerText(t, "commandCodeConnectionAdded", "Command Code connection added")
+            );
             clearCommandCodeAuthTimer();
             return;
           }
@@ -234,7 +260,11 @@ export function useCommandCodeAuth({
             setCommandCodeAuthState((current) => ({
               ...current,
               phase: "received",
-              message: "Browser approved, applying…",
+              message: providerText(
+                t,
+                "commandCodeApplyingApproval",
+                "Browser approved, applying…"
+              ),
             }));
             clearCommandCodeAuthTimer();
             await handleCommandCodeAuthApply(
@@ -258,9 +288,9 @@ export function useCommandCodeAuth({
       setCommandCodeAuthState((current) => ({
         ...current,
         phase: "error",
-        message: "Failed to start Command Code auth",
+        message: providerText(t, "commandCodeStartFailed", "Failed to start Command Code auth"),
       }));
-      notify.error("Failed to start Command Code auth");
+      notify.error(providerText(t, "commandCodeStartFailed", "Failed to start Command Code auth"));
       popup?.close?.();
       commandCodeAuthWindowRef.current = null;
       clearCommandCodeAuthTimer();
@@ -272,6 +302,7 @@ export function useCommandCodeAuth({
     fetchConnections,
     handleCommandCodeAuthApply,
     notify,
+    t,
   ]);
 
   const handleOpenCommandCodeConnect = useCallback(() => {

src/app/(dashboard)/dashboard/providers/[id]/hooks/useConnectionDeleteConfirm.ts

@@ -11,6 +11,8 @@
  */
 
 import { useCallback, useState } from "react";
+import { useTranslations } from "next-intl";
+import { providerText } from "../providerPageHelpers";
 
 export interface ConnectionDeleteConfirmTarget {
   id: string;
@@ -34,6 +36,7 @@ export function useConnectionDeleteConfirm(
   fetchConnections: () => Promise<void>,
   notify: NotifyLike
 ): ConnectionDeleteConfirmState {
+  const t = useTranslations("providers");
   const [connection, setConnection] = useState<ConnectionDeleteConfirmTarget | null>(null);
   const [deleting, setDeleting] = useState(false);
 
@@ -56,24 +59,24 @@ export function useConnectionDeleteConfirm(
     try {
       const res = await fetch(`/api/providers/${connectionId}`, { method: "DELETE" });
       if (res.ok) {
-        notify.success("Connection deleted");
+        notify.success(providerText(t, "connectionDeleted", "Connection deleted"));
         await fetchConnections();
       } else {
         const data = await res.json().catch(() => ({}));
         const message =
           (typeof data?.error === "string" && data.error) ||
           data?.error?.message ||
-          "Failed to delete connection";
+          providerText(t, "failedDeleteConnection", "Failed to delete connection");
         notify.error(message);
       }
     } catch (error) {
       console.error("Error deleting connection:", error);
-      notify.error("Failed to delete connection");
+      notify.error(providerText(t, "failedDeleteConnection", "Failed to delete connection"));
     } finally {
       setDeleting(false);
       setConnection(null);
     }
-  }, [connection, fetchConnections, notify]);
+  }, [connection, fetchConnections, notify, t]);
 
   return { connection, deleting, request, confirm, cancel };
 }

src/app/(dashboard)/dashboard/providers/[id]/hooks/useModelVisibilityHandlers.ts

@@ -316,11 +316,13 @@ export function useModelVisibilityHandlers({
         // extractApiErrorMessage coerces any object-shaped `error` (e.g. a Zod
         // format object) to a string so notify.error never hands the toast a
         // non-string child (React #31 → frozen page).
-        notify.error(extractApiErrorMessage(data, "Model test failed"));
+        notify.error(
+          extractApiErrorMessage(data, providerText(t, "modelTestFailed", "Model test failed"))
+        );
         setModelTestStatus((prev) => ({ ...prev, [modelId]: "error" }));
       }
     } catch (err) {
-      notify.error("Network error testing model");
+      notify.error(providerText(t, "modelTestNetworkError", "Network error testing model"));
       setModelTestStatus((prev) => ({ ...prev, [modelId]: "error" }));
     } finally {
       setTestingModelId(null);

src/app/(dashboard)/dashboard/providers/[id]/hooks/useProviderConnections.ts

@@ -27,7 +27,7 @@ import { useNotificationStore } from "@/store/notificationStore";
 import { isClaudeCodeCompatibleProvider } from "@/shared/constants/providers";
 import type { ConnectionRowConnection } from "../components/ConnectionRow";
 import { connectionBelongsToProviderPage } from "../../providerPageUtils";
-import { normalizeCodexLimitPolicy } from "../providerPageHelpers";
+import { normalizeCodexLimitPolicy, providerText } from "../providerPageHelpers";
 import { useProviderQuotaVisibility } from "./useProviderQuotaVisibility";
 import { useReorderByAvailability } from "./useReorderByAvailability";
 import {
@@ -378,7 +378,14 @@ export function useProviderConnections(
 
       if (!res.ok) {
         const data = await res.json().catch(() => ({}));
-        notify.error(data.error || "Failed to update Claude extra-usage policy");
+        notify.error(
+          data.error ||
+            providerText(
+              t,
+              "failedUpdateClaudeExtraUsagePolicy",
+              "Failed to update Claude extra-usage policy"
+            )
+        );
         return;
       }
 
@@ -408,12 +415,26 @@ export function useProviderConnections(
       );
       notify.success(
         enabled
-          ? "Claude extra-usage blocking enabled (extra usage will be blocked)"
-          : "Claude extra-usage blocking disabled (extra usage is allowed)"
+          ? providerText(
+              t,
+              "claudeExtraUsageBlockingEnabled",
+              "Claude extra-usage blocking enabled (extra usage will be blocked)"
+            )
+          : providerText(
+              t,
+              "claudeExtraUsageBlockingDisabled",
+              "Claude extra-usage blocking disabled (extra usage is allowed)"
+            )
       );
     } catch (error) {
       console.error("Error toggling Claude extra-usage policy:", error);
-      notify.error("Failed to update Claude extra-usage policy");
+      notify.error(
+        providerText(
+          t,
+          "failedUpdateClaudeExtraUsagePolicy",
+          "Failed to update Claude extra-usage policy"
+        )
+      );
     }
   };
 
@@ -447,7 +468,10 @@ export function useProviderConnections(
 
       if (!res.ok) {
         const data = await res.json().catch(() => ({}));
-        notify.error(data.error || "Failed to update Codex limit policy");
+        notify.error(
+          data.error ||
+            providerText(t, "failedUpdateCodexLimitPolicy", "Failed to update Codex limit policy")
+        );
         return;
       }
 
@@ -464,10 +488,12 @@ export function useProviderConnections(
             : connection
         )
       );
-      notify.success("Codex limit policy updated");
+      notify.success(providerText(t, "codexLimitPolicyUpdated", "Codex limit policy updated"));
     } catch (error) {
       console.error("Error toggling Codex quota policy:", error);
-      notify.error("Failed to update Codex limit policy");
+      notify.error(
+        providerText(t, "failedUpdateCodexLimitPolicy", "Failed to update Codex limit policy")
+      );
     }
   };
 
@@ -481,18 +507,27 @@ export function useProviderConnections(
 
       if (!res.ok) {
         const data = await res.json().catch(() => ({}));
-        notify.error(data.error || "Failed to update CLIProxyAPI routing");
+        notify.error(
+          data.error ||
+            providerText(t, "failedUpdateCliproxyRouting", "Failed to update CLIProxyAPI routing")
+        );
         return;
       }
 
       setCpaProviderEnabled(enabled);
       notify.success(
         enabled
-          ? "Requests now route through CLIProxyAPI (deeper emulation)"
-          : "Requests now use native OmniRoute (direct)"
+          ? providerText(
+              t,
+              "cliproxyRoutingEnabled",
+              "Requests now route through CLIProxyAPI (deeper emulation)"
+            )
+          : providerText(t, "cliproxyRoutingDisabled", "Requests now use native OmniRoute (direct)")
       );
     } catch {
-      notify.error("Failed to update CLIProxyAPI routing");
+      notify.error(
+        providerText(t, "failedUpdateCliproxyRouting", "Failed to update CLIProxyAPI routing")
+      );
     }
   };
 
@@ -664,10 +699,10 @@ export function useProviderConnections(
         if (onAfter) await onAfter();
       } else {
         const data = await res.json();
-        notify.error(data.error || "Batch delete failed");
+        notify.error(data.error || providerText(t, "batchDeleteFailed", "Batch delete failed"));
       }
     } catch {
-      notify.error("Network error during batch delete");
+      notify.error(providerText(t, "batchDeleteNetworkError", "Network error during batch delete"));
     } finally {
       setBatchDeleting(false);
     }
@@ -689,7 +724,11 @@ export function useProviderConnections(
         });
         if (!res.ok) {
           const data = await res.json().catch(() => ({}));
-          throw new Error(data.error?.message || data.error || "Batch update failed");
+          throw new Error(
+            data.error?.message ||
+              data.error ||
+              providerText(t, "batchUpdateFailed", "Batch update failed")
+          );
         }
         const data = await res.json();
         updated += data.updated ?? 0;
@@ -710,7 +749,10 @@ export function useProviderConnections(
         );
       }
     } catch (error: any) {
-      notify.error(error?.message || "Network error during batch update");
+      notify.error(
+        error?.message ||
+          providerText(t, "batchUpdateNetworkError", "Network error during batch update")
+      );
     } finally {
       setBatchUpdating(null);
     }
@@ -805,7 +847,13 @@ export function useProviderConnections(
       const proxiesData = await proxiesRes.json();
       const savedProxies = (proxiesData?.items || []).filter((p: any) => p.status === "active");
       if (savedProxies.length === 0) {
-        notify.error("No saved proxies found. Add proxies in Settings → Proxy first.");
+        notify.error(
+          providerText(
+            t,
+            "noSavedProxies",
+            "No saved proxies found. Add proxies in Settings → Proxy first."
+          )
+        );
         return;
       }
 
@@ -856,11 +904,16 @@ export function useProviderConnections(
       await fetchConnections();
       const tagLabel = tagFilter ? `"${tagFilter}" ` : "";
       notify.success(
-        `Distributed ${assigned} proxy assignment(s) across ${tagLabel}${sorted.length} connection(s).`
+        providerText(
+          t,
+          "proxiesDistributed",
+          "Distributed {assigned} proxy assignment(s) across {tagLabel}{total} connection(s).",
+          { assigned, tagLabel, total: sorted.length }
+        )
       );
     } catch (err) {
       console.error("Error distributing proxies:", err);
-      notify.error("Failed to distribute proxies.");
+      notify.error(providerText(t, "failedDistributeProxies", "Failed to distribute proxies."));
     } finally {
       setDistributingProxies(false);
     }

src/app/(dashboard)/dashboard/providers/[id]/hooks/useProviderModels.ts

@@ -16,7 +16,7 @@
 import { useState, useCallback } from "react";
 import { useTranslations } from "next-intl";
 import { useNotificationStore } from "@/store/notificationStore";
-import type { CompatModelRow } from "../providerPageHelpers";
+import { providerText, type CompatModelRow } from "../providerPageHelpers";
 
 // ──── types ─────────────────────────────────────────────────────────────────
 
@@ -79,11 +79,13 @@ export function useProviderModels(
           notify.success(t("setAliasSuccess", { alias }));
         } else {
           const data = await res.json().catch(() => ({}));
-          notify.error(data?.error?.message || "Failed to set alias");
+          notify.error(
+            data?.error?.message || providerText(t, "failedSetAlias", "Failed to set alias")
+          );
         }
       } catch (error) {
         console.log("Error setting alias:", error);
-        notify.error("Network error setting alias");
+        notify.error(providerText(t, "networkErrorSettingAlias", "Network error setting alias"));
       }
     },
     [fetchAliases, t, notify]
@@ -100,11 +102,13 @@ export function useProviderModels(
           notify.success(t("deleteAliasSuccess", { alias }));
         } else {
           const data = await res.json().catch(() => ({}));
-          notify.error(data?.error?.message || "Failed to delete alias");
+          notify.error(
+            data?.error?.message || providerText(t, "failedDeleteAlias", "Failed to delete alias")
+          );
         }
       } catch (error) {
         console.log("Error deleting alias:", error);
-        notify.error("Network error deleting alias");
+        notify.error(providerText(t, "networkErrorDeletingAlias", "Network error deleting alias"));
       }
     },
     [fetchAliases, t, notify]
@@ -113,10 +117,9 @@ export function useProviderModels(
   const fetchProviderModelMeta = useCallback(async () => {
     if (isSearchProvider) return;
     try {
-      const res = await fetch(
-        `/api/provider-models?provider=${encodeURIComponent(providerId)}`,
-        { cache: "no-store" }
-      );
+      const res = await fetch(`/api/provider-models?provider=${encodeURIComponent(providerId)}`, {
+        cache: "no-store",
+      });
       if (!res.ok) return;
       const data = await res.json();
       setModelMeta({

src/app/(dashboard)/dashboard/providers/[id]/hooks/useProviderSettings.ts

@@ -115,9 +115,7 @@ export function useProviderSettings(providerId: string): UseProviderSettingsRetu
     } catch (error) {
       if (!isCurrentRequest()) return;
       setCodexSettingsLoaded(false);
-      setCodexSettingsLoadError(
-        error instanceof Error ? error.message : "Failed to load settings"
-      );
+      setCodexSettingsLoadError(error instanceof Error ? error.message : "Failed to load settings");
     }
   }, [providerId]);
 
@@ -184,15 +182,20 @@ export function useProviderSettings(providerId: string): UseProviderSettingsRetu
       if (!res.ok) {
         const data = await res.json().catch(() => ({}));
         setCodexGlobalServiceMode(previousMode);
-        notify.error(data.error || "Failed to update Codex service mode");
+        notify.error(
+          data.error ||
+            providerText(t, "failedUpdateCodexServiceMode", "Failed to update Codex service mode")
+        );
         return;
       }
 
-      notify.success("Codex service mode updated");
+      notify.success(providerText(t, "codexServiceModeUpdated", "Codex service mode updated"));
     } catch (error) {
       setCodexGlobalServiceMode(previousMode);
       console.error("Error updating Codex service mode:", error);
-      notify.error("Failed to update Codex service mode");
+      notify.error(
+        providerText(t, "failedUpdateCodexServiceMode", "Failed to update Codex service mode")
+      );
     } finally {
       setSavingCodexGlobalServiceMode(false);
     }
@@ -215,7 +218,14 @@ export function useProviderSettings(providerId: string): UseProviderSettingsRetu
       if (!res.ok) {
         const data = await res.json().catch(() => ({}));
         setPreferClaudeCodeForUnprefixedClaudeModels(previous);
-        notify.error(data.error || "Failed to update Claude Code routing preference");
+        notify.error(
+          data.error ||
+            providerText(
+              t,
+              "failedUpdateClaudeRoutingPreference",
+              "Failed to update Claude Code routing preference"
+            )
+        );
         return;
       }
 
@@ -227,14 +237,26 @@ export function useProviderSettings(providerId: string): UseProviderSettingsRetu
       }
       notify.success(
         enabled
-          ? "Unprefixed Claude models now prefer Claude Code"
-          : "Unprefixed Claude models no longer prefer Claude Code"
+          ? providerText(
+              t,
+              "claudeRoutingPreferenceEnabled",
+              "Unprefixed Claude models now prefer Claude Code"
+            )
+          : providerText(
+              t,
+              "claudeRoutingPreferenceDisabled",
+              "Unprefixed Claude models no longer prefer Claude Code"
+            )
       );
     } catch (error) {
       setPreferClaudeCodeForUnprefixedClaudeModels(previous);
       console.error("Error updating Claude Code routing preference:", error);
       notify.error(
-        providerText(t, "failedUpdateClaudeRoutingPreference", "Failed to update Claude Code routing preference")
+        providerText(
+          t,
+          "failedUpdateClaudeRoutingPreference",
+          "Failed to update Claude Code routing preference"
+        )
       );
     } finally {
       setSavingClaudeRoutingPreference(false);

src/app/(dashboard)/dashboard/providers/components/AddCompatibleProviderModal.tsx

@@ -8,6 +8,7 @@ import {
   CLIENT_IDENTITY_PROFILE_OPTIONS,
   getClientIdentityProfileHeaders,
 } from "@/shared/constants/clientIdentityProfiles";
+import { providerText } from "../[id]/providerPageHelpers";
 
 type CompatibleMode = "openai" | "anthropic" | "cc";
 type CompatibleProviderNode = { id: string } & Record<string, unknown>;
@@ -100,9 +101,11 @@ export default function AddCompatibleProviderModal({
   const [checkKey, setCheckKey] = useState("");
   const [checkModelId, setCheckModelId] = useState("");
   const [validating, setValidating] = useState(false);
-  const [validationResult, setValidationResult] = useState<
-    null | { valid: boolean; error?: string | null; method?: string | null }
-  >(null);
+  const [validationResult, setValidationResult] = useState<null | {
+    valid: boolean;
+    error?: string | null;
+    method?: string | null;
+  }>(null);
   const [showAdvanced, setShowAdvanced] = useState(false);
 
   const apiTypeOptions = useMemo(
@@ -242,7 +245,10 @@ export default function AddCompatibleProviderModal({
         method: data.method ?? null,
       });
     } catch {
-      setValidationResult({ valid: false, error: "Network error" });
+      setValidationResult({
+        valid: false,
+        error: providerText(t, "networkError", "Network error"),
+      });
     } finally {
       setValidating(false);
     }

src/app/(dashboard)/dashboard/providers/components/ProviderCard.tsx

@@ -16,7 +16,7 @@ import {
 } from "@/shared/constants/providers";
 
 import { CategoryDot } from "./CategoryDot";
-import { isKimiPartnerProviderId } from "../featuredProviders";
+import { isCheaperInferenceProviderId, isKimiPartnerProviderId } from "../featuredProviders";
 
 interface ProviderStats {
   total?: number;
@@ -30,17 +30,17 @@ interface ProviderStats {
   codexServiceTier?: "default" | "priority" | "flex" | null;
 }
 
-const KIND_LABEL: Record<string, string> = {
-  llm: "Chat",
-  embedding: "Embed",
-  image: "Image",
-  imageToText: "I→T",
-  tts: "TTS",
-  stt: "STT",
-  webSearch: "Search",
-  webFetch: "Fetch",
-  video: "Video",
-  music: "Music",
+const KIND_LABEL_KEYS: Record<string, { key: string; fallback: string }> = {
+  llm: { key: "serviceKindChat", fallback: "Chat" },
+  embedding: { key: "serviceKindEmbedding", fallback: "Embed" },
+  image: { key: "serviceKindImage", fallback: "Image" },
+  imageToText: { key: "serviceKindImageToText", fallback: "I→T" },
+  tts: { key: "serviceKindTts", fallback: "TTS" },
+  stt: { key: "serviceKindStt", fallback: "STT" },
+  webSearch: { key: "serviceKindWebSearch", fallback: "Search" },
+  webFetch: { key: "serviceKindWebFetch", fallback: "Fetch" },
+  video: { key: "serviceKindVideo", fallback: "Video" },
+  music: { key: "serviceKindMusic", fallback: "Music" },
 };
 
 /** Maps a compatible-provider `apiType` to its `KIND_LABEL` key (#6936: non-chat
@@ -166,6 +166,10 @@ const ProviderCard = forwardRef<ProviderCardHandle, ProviderCardProps>(function
   const t = useTranslations("providers");
   const tc = useTranslations("common");
   const tp = useTranslations("miniPlayground");
+  const kindLabel = (kind: string) => {
+    const entry = KIND_LABEL_KEYS[kind];
+    return entry ? providerText(t, entry.key, entry.fallback) : kind;
+  };
   const [testExpanded, setTestExpanded] = useState<boolean>(false);
   const innerRef = useRef<HTMLDivElement>(null);
   const linkElementRef = useRef<HTMLAnchorElement>(null);
@@ -223,9 +227,11 @@ const ProviderCard = forwardRef<ProviderCardHandle, ProviderCardProps>(function
   const isCompatible = isOpenAICompatibleProvider(providerId);
   const isCcCompatible = isClaudeCodeCompatibleProvider(providerId);
   const isAnthropicCompatible = isAnthropicCompatibleProvider(providerId) && !isCcCompatible;
-  // Kimi (Moonshot AI) official-partnership highlight (2026-07): UI-only accent,
-  // see featuredProviders.ts — never affects routing/fallback order.
+  // Open Source Friend highlights (Kimi 2026-07, Cheaper Inference 2026-07): UI-only
+  // accents, see featuredProviders.ts — never affect routing/fallback order.
   const isKimiPartner = isKimiPartnerProviderId(provider.id || providerId);
+  const isCheaperInferencePartner = isCheaperInferenceProviderId(provider.id || providerId);
+  const isSponsorPartner = isKimiPartner || isCheaperInferencePartner;
   const codexServiceTierLabel =
     stats.codexServiceTier === "flex"
       ? providerText(t, "codexTierFlexLabel", "Flex")
@@ -267,6 +273,24 @@ const ProviderCard = forwardRef<ProviderCardHandle, ProviderCardProps>(function
     </span>
   ) : null;
 
+  // Cheaper Inference Open Source Friend badge — brand green (#31f889) with a dark
+  // foreground (the green is too bright for white text). Literal Tailwind arbitrary
+  // values must stay in sync with CHEAPERINFERENCE_BRAND_COLOR (featuredProviders.ts).
+  const cheaperInferenceSupporterChip = isCheaperInferencePartner ? (
+    <span
+      key="cheaperinference-supporter"
+      className="inline-flex items-center gap-0.5 rounded-full border border-[#31f889]/40 bg-[#31f889]/15 px-1.5 py-0 text-[9px] font-semibold uppercase tracking-wide leading-none text-[#0b7a45] dark:text-[#5CF0A6]"
+      title={providerText(
+        t,
+        "cheaperInferenceSupporterTooltip",
+        "Cheaper Inference backs OmniRoute as an Open Source Friend"
+      )}
+    >
+      <span className="material-symbols-outlined text-[10px] leading-none">verified</span>
+      {providerText(t, "cheaperInferenceSupporterBadge", "Open Source Friend")}
+    </span>
+  ) : null;
+
   const dotLabels: Record<string, string> = {
     free: tc("free"),
     "no-auth": t("noAuthLabel"),
@@ -318,7 +342,12 @@ const ProviderCard = forwardRef<ProviderCardHandle, ProviderCardProps>(function
                 // is a raw (non-token) brand hex, not a theme color. Keep the hex in
                 // sync with KIMI_BRAND_COLOR (featuredProviders.ts).
                 "border-2 border-[#1783FF]/70 hover:border-[#1783FF]/90 shadow-[inset_0_0_0_100px_rgba(23,131,255,0.035),0_4px_16px_-4px_rgba(23,131,255,0.45)]"
-              : "hover:border-primary/40"
+              : isCheaperInferencePartner
+                ? // Cheaper Inference Open Source Friend accent — same construction in
+                  // its brand green (#31f889 = rgb(49,248,137)). Keep in sync with
+                  // CHEAPERINFERENCE_BRAND_COLOR (featuredProviders.ts).
+                  "border-2 border-[#31f889]/70 hover:border-[#31f889]/90 shadow-[inset_0_0_0_100px_rgba(49,248,137,0.035),0_4px_16px_-4px_rgba(49,248,137,0.45)]"
+                : "hover:border-primary/40"
           } ${allDisabled ? "opacity-50" : ""} ${provider.deprecated ? "opacity-60" : ""}`}
         >
           <div className="flex flex-col gap-2 h-full">
@@ -392,23 +421,24 @@ const ProviderCard = forwardRef<ProviderCardHandle, ProviderCardProps>(function
               isCompatible ||
               isCcCompatible ||
               isAnthropicCompatible ||
-              isKimiPartner) && (
+              isSponsorPartner) && (
               <div className="flex flex-wrap items-center gap-1">
                 {kimiOfficialSupporterChip}
+                {cheaperInferenceSupporterChip}
                 {provider.serviceKinds?.map((k) => (
                   <span
                     key={k}
                     className="text-[10px] px-1.5 py-0.5 rounded bg-bg-subtle border border-border text-text-muted leading-none"
                   >
-                    {KIND_LABEL[k] ?? k}
+                    {kindLabel(k)}
                   </span>
                 ))}
                 {isCompatible && (
                   <Badge variant="default" size="sm">
                     {provider.apiType === "responses"
                       ? t("responses")
-                      : (KIND_LABEL[COMPATIBLE_API_TYPE_KIND[provider.apiType ?? ""] ?? ""] ??
-                        t("chat"))}
+                      : kindLabel(COMPATIBLE_API_TYPE_KIND[provider.apiType ?? ""] ?? "") ||
+                        t("chat")}
                   </Badge>
                 )}
                 {isCcCompatible && (