fix(ci): stamp dist/BUILD_SHA before the npm publish provenance gate

The publish job builds with `build:cli`, which assembles dist/ but does not write dist/BUILD_SHA — only `build:release` does, via write-build-sha.mjs. The #10427 provenance guard inside check:pack-artifact then rejects the artifact for having no SHA, so the build+validate pair in this job could never pass: [provenance] dist/BUILD_SHA is missing — the artifact cannot be traced to a commit. This is the same structural gap that was fixed in ci.yml's Package Artifact job earlier in the v3.8.50 cycle; npm-publish.yml carried it too and it only became visible now that the job finally got past the runner's disk exhaustion. Stamp from github.sha (on a release event that is the tag commit, which is on main) and fetch origin/main so the ancestry probe can resolve the ref that the guard checks against by default.
2026-09-01 12:22:24 +03:00 · 2026-08-26 20:39:33 -03:00
1358 changed files with 2496 additions and 4895 deletions
--- a/.env.example
+++ b/.env.example
@@ -1023,13 +1023,6 @@ PROVIDER_LIMITS_SYNC_SPACING_MS=1500
 # to disable the check. Used by: src/lib/db/migrationRunner.ts.  Default: 50.
 #OMNIROUTE_MAX_PENDING_MIGRATIONS=50

-# Working directory for the check:install-upgrade release gate. It builds two ~3 GB
-# install trees plus a ~275 MB tarball, so it needs roughly 12 GB — more than the
-# 12 GB RAM-backed tmpfs that /tmp is on the self-hosted runner, where it exhausted
-# the tmpfs and npm silently truncated the package. Defaults to <repo>/.install-upgrade
-# on real disk. Used by: scripts/check/check-install-upgrade.mjs.  Default: <repo>/.install-upgrade.
-#OMNIROUTE_INSTALL_UPGRADE_WORKDIR=/var/tmp/omniroute-install-upgrade
-
 # Trust user-managed RTK project filter rules without strict signature checks.
 # Used by: open-sse/services/compression/engines/rtk/filterLoader.ts.  Default: 0.
 #OMNIROUTE_RTK_TRUST_PROJECT_FILTERS=0
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -606,24 +606,13 @@ jobs:
    # Dynamic runner: when the release captain flips the USE_VPS_RUNNER repo var to
    # 'true' (scripts/vps/release-runner-up.sh does it after the self-hosted VM is
    # online), the heavy jobs run on the dedicated 32-core VPS runners (label
-    # omni-build) instead of queueing on the 20-concurrent-job hosted pool.
+    # omni-release) instead of queueing on the 20-concurrent-job hosted pool.
    # Safety: fork PRs NEVER reach the self-hosted runner — the expression falls
    # back to ubuntu-latest unless the PR head repo is this repository (push /
    # dispatch events are own-origin by definition). Any failure path (VM down,
    # var unset/false) also falls back to ubuntu-latest.
-    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && (github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name == github.repository)) && fromJSON('["self-hosted","omni-build"]') || 'ubuntu-latest' }}
+    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && (github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name == github.repository)) && fromJSON('["self-hosted","omni-release"]') || 'ubuntu-latest' }}
    needs: changes
-    # The .113 pool runs ONE next-build with room to spare and two at the edge: the
-    # box has 31 GB and a single next-build peaks at 14–16 GB RSS. On 2026-08-28
-    # 13:50Z the kernel OOM-killed main's build while a PR build ran beside it
-    # (five Build jobs had been queued by a burst of PRs). Two lanes: main keeps
-    # its own so a release is never queued behind PR traffic; PR builds serialize
-    # among themselves. GitHub keeps one running + one pending per group and
-    # CANCELS older pendings — a cancelled PR build is re-runnable; a dead main
-    # build costs the publish its artefact and a 40-minute rebuild that OOMs.
-    concurrency:
-      group: heavy-build-${{ github.ref == 'refs/heads/main' && 'main' || 'pr' }}
-      cancel-in-progress: false
    if: ${{ github.event_name != 'pull_request' || (needs.changes.outputs.code == 'true' && github.event.pull_request.draft == false) }}
    steps:
      - uses: actions/checkout@v7
@@ -657,14 +646,14 @@ jobs:
        # Keep standalone/node_modules intact: package/electron jobs consume the
        # Next-traced standalone tree and must not replace it with root node_modules.
        run: |
-          tar -czf "$RUNNER_TEMP/e2e-build.tar.gz" \
+          tar -czf /tmp/e2e-build.tar.gz \
            --exclude='.build/next/cache' \
            .build/next
      - name: Upload Next.js build for downstream jobs
        uses: actions/upload-artifact@v7
        with:
          name: next-build
-          path: ${{ runner.temp }}/e2e-build.tar.gz
+          path: /tmp/e2e-build.tar.gz
          retention-days: 1

  package-artifact:
@@ -687,14 +676,10 @@ jobs:
        uses: actions/download-artifact@v8
        with:
          name: next-build
-          # Workspace-relative on purpose: the matrix below includes windows-latest, whose
-          # default shell is pwsh, where $RUNNER_TEMP is empty (it is $env:RUNNER_TEMP) —
-          # #11896's first cut broke the Electron smoke on exactly that. A relative path
-          # works in bash and pwsh alike; hosted workspaces are ephemeral.
-          path: next-build-artifact
+          path: /tmp/
      - name: Extract Next.js build artifact
        run: |
-          tar -xzf next-build-artifact/e2e-build.tar.gz
+          tar -xzf /tmp/e2e-build.tar.gz
      # build:cli consumes the downloaded .build/next standalone artifact and assembles dist/;
      # it only rebuilds if the downloaded standalone artifact is missing.
      - run: npm run build:cli
@@ -782,14 +767,10 @@ jobs:
        uses: actions/download-artifact@v8
        with:
          name: next-build
-          # Workspace-relative on purpose: the matrix below includes windows-latest, whose
-          # default shell is pwsh, where $RUNNER_TEMP is empty (it is $env:RUNNER_TEMP) —
-          # #11896's first cut broke the Electron smoke on exactly that. A relative path
-          # works in bash and pwsh alike; hosted workspaces are ephemeral.
-          path: next-build-artifact
+          path: /tmp/
      - name: Extract Next.js build artifact
        run: |
-          tar -xzf next-build-artifact/e2e-build.tar.gz
+          tar -xzf /tmp/e2e-build.tar.gz
      - name: Install Electron dependencies
        working-directory: electron
        run: npm install --no-audit --no-fund
@@ -976,11 +957,7 @@ jobs:
    # 10min was sized before #7114 added the lcov reporter (Codecov/Sonar need it);
    # merging 8 shard JSONs + text+json+lcov now takes ~10-12min — three consecutive
    # release-tip runs died at exactly 10m as job-timeout "cancelled" (2026-07-15/16).
-    # 30, not 20 (2026-08-29): the informational Codecov upload below hung for the rest of
-    # the budget on two consecutive main runs (33207760653, 33215115341); the job ended
-    # `cancelled` and dragged the whole run's conclusion to `cancelled` although every
-    # blocking job was green. The upload step now has its own ceiling; this is headroom.
-    timeout-minutes: 30
+    timeout-minutes: 20
    needs: test-unit
    if: ${{ !cancelled() && needs.test-unit.result == 'success' && !contains(github.event.pull_request.labels.*.name, 'hotfix') }}
    env:
@@ -1059,10 +1036,6 @@ jobs:
      # (if-no-files-found: warn) — Sonar consumes the same file.
      - name: Upload coverage to Codecov (informational)
        if: always()
-        # Informational means informational: its own ceiling and continue-on-error, so a
-        # stalled upload can neither eat the job's budget nor turn a green job cancelled.
-        timeout-minutes: 5
-        continue-on-error: true
        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
        with:
          files: coverage/lcov.info
@@ -1257,14 +1230,10 @@ jobs:
        uses: actions/download-artifact@v8
        with:
          name: next-build
-          # Workspace-relative on purpose: the matrix below includes windows-latest, whose
-          # default shell is pwsh, where $RUNNER_TEMP is empty (it is $env:RUNNER_TEMP) —
-          # #11896's first cut broke the Electron smoke on exactly that. A relative path
-          # works in bash and pwsh alike; hosted workspaces are ephemeral.
-          path: next-build-artifact
+          path: /tmp/
      - name: Extract Next.js build artifact
        run: |
-          tar -xzf next-build-artifact/e2e-build.tar.gz
+          tar -xzf /tmp/e2e-build.tar.gz
      # WS4.1: duration-balanced shards (LPT over config/quality/e2e-timings.json).
      # Measured skew of plain --shard was 14× (24m47s vs 1m47s) — E2E was the CI
      # critical path. The balancer self-verifies completeness and exits non-zero on
--- a/.github/workflows/docker-publish.yml
+++ b/.github/workflows/docker-publish.yml
@@ -185,13 +185,6 @@ jobs:

      - name: Build and push BUN base platform image by digest
        id: build-bun-base
-        # Bun is a best-effort compatibility target, not a supported runtime
-        # (AGENTS.md -> Environment). Its `bun run build` has been OOM-killing on
-        # both arches; letting that sink the whole publish means the SUPPORTED
-        # runner-base / runner-web images never reach the registry either. The
-        # image is still built and pushed whenever it succeeds — only its power to
-        # block the release is removed.
-        continue-on-error: true
        uses: docker/build-push-action@53b7df96c91f9c12dcc8a07bcb9ccacbed38856a # v7
        with:
          context: .
@@ -210,13 +203,6 @@ jobs:

      - name: Build and push BUN web platform image by digest
        id: build-bun-web
-        # Bun is a best-effort compatibility target, not a supported runtime
-        # (AGENTS.md -> Environment). Its `bun run build` has been OOM-killing on
-        # both arches; letting that sink the whole publish means the SUPPORTED
-        # runner-base / runner-web images never reach the registry either. The
-        # image is still built and pushed whenever it succeeds — only its power to
-        # block the release is removed.
-        continue-on-error: true
        uses: docker/build-push-action@53b7df96c91f9c12dcc8a07bcb9ccacbed38856a # v7
        with:
          context: .
@@ -244,15 +230,8 @@ jobs:
          mkdir -p /tmp/digests/base /tmp/digests/web /tmp/digests/bun-base /tmp/digests/bun-web
          touch "/tmp/digests/base/${DIGEST_BASE#sha256:}"
          touch "/tmp/digests/web/${DIGEST_WEB#sha256:}"
-          # Empty when the (non-blocking) bun build produced no image. `if` blocks,
-          # not `[ -n ] && touch`: under `set -e` a failing AND-list aborts the step,
-          # which is precisely the case being handled here.
-          if [ -n "$DIGEST_BUN_BASE" ]; then
-            touch "/tmp/digests/bun-base/${DIGEST_BUN_BASE#sha256:}"
-          fi
-          if [ -n "$DIGEST_BUN_WEB" ]; then
-            touch "/tmp/digests/bun-web/${DIGEST_BUN_WEB#sha256:}"
-          fi
+          touch "/tmp/digests/bun-base/${DIGEST_BUN_BASE#sha256:}"
+          touch "/tmp/digests/bun-web/${DIGEST_BUN_WEB#sha256:}"

      - name: Upload base digests
        uses: actions/upload-artifact@v7
@@ -275,11 +254,7 @@ jobs:
        with:
          name: digests-bun-base-${{ matrix.arch }}
          path: /tmp/digests/bun-base/*
-          # `ignore`, not `error`: the bun build is non-blocking, so an absent
-          # digest is the expected outcome of a failed/skipped bun image — the
-          # manifest step already treats these tags as optional. Leaving `error`
-          # here just relocates the blocker from the manifest to the upload.
-          if-no-files-found: ignore
+          if-no-files-found: error
          retention-days: 1

      - name: Upload bun-web digests
@@ -287,11 +262,7 @@ jobs:
        with:
          name: digests-bun-web-${{ matrix.arch }}
          path: /tmp/digests/bun-web/*
-          # `ignore`, not `error`: the bun build is non-blocking, so an absent
-          # digest is the expected outcome of a failed/skipped bun image — the
-          # manifest step already treats these tags as optional. Leaving `error`
-          # here just relocates the blocker from the manifest to the upload.
-          if-no-files-found: ignore
+          if-no-files-found: error
          retention-days: 1

  merge:
@@ -349,9 +320,6 @@ jobs:
          merge-multiple: true

      - name: Download bun-base digests
-        # Non-blocking: the bun image is best-effort, so its artifact may not
-        # exist at all. The manifest step treats these tags as optional.
-        continue-on-error: true
        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8
        with:
          pattern: digests-bun-base-*
@@ -359,9 +327,6 @@ jobs:
          merge-multiple: true

      - name: Download bun-web digests
-        # Non-blocking: the bun image is best-effort, so its artifact may not
-        # exist at all. The manifest step treats these tags as optional.
-        continue-on-error: true
        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8
        with:
          pattern: digests-bun-web-*
@@ -373,7 +338,7 @@ jobs:
          set -euo pipefail

          create_manifest() {
-            local image="$1" suffix="$2" dir="$3" optional="${4:-}"
+            local image="$1" suffix="$2" dir="$3"
            local tags=(-t "${image}:${VERSION}${suffix}")
            if [ "$PROMOTE_LATEST" = "true" ]; then
              tags+=(-t "${image}:latest${suffix}")
@@ -383,10 +348,6 @@ jobs:
              refs+=("${image}@sha256:$(basename "$digest_file")")
            done < <(find "$dir" -type f | sort)
            if [ "${#refs[@]}" -eq 0 ]; then
-              if [ -n "$optional" ]; then
-                echo "::warning::No image digests in $dir — skipping optional tag ${image}:${VERSION}${suffix}" >&2
-                return 0
-              fi
              echo "No image digests in $dir" >&2
              exit 1
            fi
@@ -395,15 +356,15 @@ jobs:

          create_manifest "${IMAGE_NAME}" "" /tmp/digests/base
          create_manifest "${IMAGE_NAME}" "-web" /tmp/digests/web
-          create_manifest "${IMAGE_NAME}" "-bun" /tmp/digests/bun-base optional
-          create_manifest "${IMAGE_NAME}" "-web-bun" /tmp/digests/bun-web optional
+          create_manifest "${IMAGE_NAME}" "-bun" /tmp/digests/bun-base
+          create_manifest "${IMAGE_NAME}" "-web-bun" /tmp/digests/bun-web

      - name: Create GHCR manifest
        run: |
          set -euo pipefail

          create_manifest() {
-            local image="$1" suffix="$2" dir="$3" optional="${4:-}"
+            local image="$1" suffix="$2" dir="$3"
            local tags=(-t "${image}:${VERSION}${suffix}")
            if [ "$PROMOTE_LATEST" = "true" ]; then
              tags+=(-t "${image}:latest${suffix}")
@@ -413,10 +374,6 @@ jobs:
              refs+=("${image}@sha256:$(basename "$digest_file")")
            done < <(find "$dir" -type f | sort)
            if [ "${#refs[@]}" -eq 0 ]; then
-              if [ -n "$optional" ]; then
-                echo "::warning::No image digests in $dir — skipping optional tag ${image}:${VERSION}${suffix}" >&2
-                return 0
-              fi
              echo "No image digests in $dir" >&2
              exit 1
            fi
@@ -425,8 +382,8 @@ jobs:

          create_manifest "${GHCR_IMAGE_NAME}" "" /tmp/digests/base
          create_manifest "${GHCR_IMAGE_NAME}" "-web" /tmp/digests/web
-          create_manifest "${GHCR_IMAGE_NAME}" "-bun" /tmp/digests/bun-base optional
-          create_manifest "${GHCR_IMAGE_NAME}" "-web-bun" /tmp/digests/bun-web optional
+          create_manifest "${GHCR_IMAGE_NAME}" "-bun" /tmp/digests/bun-base
+          create_manifest "${GHCR_IMAGE_NAME}" "-web-bun" /tmp/digests/bun-web

      - name: Inspect image
        if: needs.prepare.outputs.version != 'main'
@@ -485,9 +442,6 @@ jobs:
          severity: CRITICAL
          ignore-unfixed: true
          exit-code: "1"
-          # Explicit: the advisory scan above already points at it, and the blocking
-          # gate must honour the same accepted-risk list (#12084).
-          trivyignores: .trivyignore

      - name: Upload Trivy SARIF to Security tab
        if: needs.prepare.outputs.version != 'main'
--- a/.github/workflows/electron-release.yml
+++ b/.github/workflows/electron-release.yml
@@ -4,21 +4,12 @@ on:
  push:
    tags:
      - "v*"
-  # A dispatch builds the ref it is dispatched ON (`gh workflow run … --ref v3.8.50` rebuilds
-  # that tag; `--ref main` builds the repaired line). The ref is deliberately NOT an input:
-  # CodeQL flags an input-controlled checkout next to the npm cache on the default branch as
-  # cache poisoning (actions/cache-poisoning/poisonable-step), and `github.ref` is trusted.
  workflow_dispatch:
    inputs:
      version:
        description: "Release version (e.g., v1.6.8)"
        required: true
        type: string
-      publish_npm:
-        description: "Also run the npm publish leg (turn off when re-attaching desktop assets to a release whose npm package already shipped)"
-        required: false
-        default: true
-        type: boolean

 # Least-privilege default: read-only at the top level; each job grants the writes it
 # needs (build/release upload assets, publish-npm forwards npm provenance / packages
@@ -413,14 +404,7 @@ jobs:
          tag_name: ${{ needs.validate.outputs.version }}
          draft: false
          prerelease: false
-          # NEVER. Phase 3 of the release flow creates the GitHub Release with the curated
-          # notes seconds after pushing the tag, so by the time this step runs (1-2 h of
-          # builds later) the body already exists — and `true` APPENDS GitHub's
-          # auto-generated "What's Changed" block to it (v3.8.48 shipped that way; the
-          # v3.8.50 re-attach dispatch added +1,416 chars to a 121 KB body, run
-          # 33238093090). A curated body sits ~3 KB under the 125,000-char cap, so the
-          # append can also turn this step RED and leave the release with no assets.
-          generate_release_notes: false
+          generate_release_notes: true
          fail_on_unmatched_files: false
          files: |
            release-assets/*.dmg
@@ -478,20 +462,11 @@ jobs:
  publish-npm:
    name: Publish to npm
    needs: [validate, release]
-    # A re-dispatch that only re-attaches desktop assets must not publish the npm package again.
-    if: ${{ github.event_name != 'workflow_dispatch' || inputs.publish_npm }}
    permissions:
      # Must be `write`, not `read`: this job calls the reusable npm-publish.yml whose
      # `publish` job needs `contents: write` (gh release upload — attach the SBOM, #3874).
      # A reusable workflow's job cannot request more permission than the caller grants,
      # so a `read` here makes GitHub reject the run at startup (startup_failure).
-      #
-      # `actions: read` for the same reason: the called `publish` job downloads the next-build
-      # artefact and requests it. v3.8.50 (run 33005490476) died at startup with "The nested
-      # job 'publish' is requesting 'actions: read', but is only allowed 'actions: none'" — and
-      # because `release` lives in this same workflow, the tag shipped with ZERO assets. Keep
-      # this block a superset of every job's permissions in npm-publish.yml.
-      actions: read
      contents: write
      id-token: write # npm provenance (forwarded to the reusable workflow)
      packages: write # publish to npm.pkg.github.com
--- a/.github/workflows/nightly-release-green.yml
+++ b/.github/workflows/nightly-release-green.yml
@@ -68,7 +68,7 @@ jobs:
    # this runs on the dedicated VPS runner — clean env (no operator OMNIROUTE_API_KEY,
    # no local noauth CLIs => zero machine-specific false positives) and no contention.
    # Nightly cron normally finds the var false (VM off) and falls back to hosted.
-    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && fromJSON('["self-hosted","omni-build"]')) || 'ubuntu-latest' }}
+    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && fromJSON('["self-hosted","omni-release"]')) || 'ubuntu-latest' }}
    env:
      JWT_SECRET: ci-nightly-secret-with-sufficient-length-for-validation
      API_KEY_SECRET: ci-nightly-api-key-secret-long
@@ -196,26 +196,6 @@ jobs:
            gh issue create --repo "$GITHUB_REPOSITORY" --title "$TITLE" --label base-red --body-file issue-body.md
          fi

-      - name: Close tracking issue when the branch is green again
-        if: steps.validate.outputs.exit == '0'
-        env:
-          GH_TOKEN: ${{ github.token }}
-          TARGET: ${{ steps.branch.outputs.target }}
-          RUN_URL: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}
-        run: |
-          set -euo pipefail
-          # The open/update step above is the UPWARD half of the loop; without this
-          # step a stale "not green" issue outlives the fix and every base-green check
-          # (`AGENTS.md` → "Base-green check") keeps stamping new PRs as base-red inherited.
-          TITLE="🔴 Release branch not green: ${TARGET}"
-          EXISTING=$(gh issue list --repo "$GITHUB_REPOSITORY" --state open \
-            --search "in:title $TITLE" --json number --jq '.[0].number' 2>/dev/null || echo "")
-          if [ -n "$EXISTING" ]; then
-            gh issue close "$EXISTING" --repo "$GITHUB_REPOSITORY" --reason completed \
-              --comment "✅ \`${TARGET}\` is release-green again at \`${GITHUB_SHA:0:9}\` — ${RUN_URL}. Auto-closed by Release-Green (continuous)."
-            echo "Closed issue #$EXISTING"
-          fi
-
      - name: Upload report artifact
        if: always()
        uses: actions/upload-artifact@v7
@@ -237,7 +217,7 @@ jobs:
    # On a push, only run for a push to main — a push to release/* is handled by
    # release-green above. Schedule/dispatch always run (they also sweep main).
    if: ${{ github.event_name != 'push' || github.ref_name == 'main' }}
-    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && fromJSON('["self-hosted","omni-build"]')) || 'ubuntu-latest' }}
+    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && fromJSON('["self-hosted","omni-release"]')) || 'ubuntu-latest' }}
    env:
      JWT_SECRET: ci-nightly-secret-with-sufficient-length-for-validation
      API_KEY_SECRET: ci-nightly-api-key-secret-long
@@ -314,25 +294,6 @@ jobs:
            gh issue create --repo "$GITHUB_REPOSITORY" --title "$TITLE" --label base-red --body-file issue-body.md
          fi

-      - name: Close tracking issue when the branch is green again
-        if: steps.validate.outputs.exit == '0'
-        env:
-          GH_TOKEN: ${{ github.token }}
-          RUN_URL: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}
-        run: |
-          set -euo pipefail
-          # The open/update step above is the UPWARD half of the loop; without this
-          # step a stale "not green" issue outlives the fix and every base-green check
-          # (`AGENTS.md` → "Base-green check") keeps stamping new PRs as base-red inherited.
-          TITLE="🔴 main branch not green"
-          EXISTING=$(gh issue list --repo "$GITHUB_REPOSITORY" --state open \
-            --search "in:title $TITLE" --json number --jq '.[0].number' 2>/dev/null || echo "")
-          if [ -n "$EXISTING" ]; then
-            gh issue close "$EXISTING" --repo "$GITHUB_REPOSITORY" --reason completed \
-              --comment "✅ \`main\` is main-green again at \`${GITHUB_SHA:0:9}\` — ${RUN_URL}. Auto-closed by Release-Green (continuous)."
-            echo "Closed issue #$EXISTING"
-          fi
-
      - name: Upload report artifact
        if: always()
        uses: actions/upload-artifact@v7
--- a/.github/workflows/npm-publish.yml
+++ b/.github/workflows/npm-publish.yml
@@ -23,12 +23,11 @@ on:
          - next
          - historic
      publish_mode:
-        description: "auto = publish through npm Trusted Publishing (OIDC, no token, no 2FA prompt — the default); staged = npm stage publish (owner approves with 2FA); direct = legacy token publish (emergency fallback only)"
+        description: "staged = npm stage publish (owner approves with 2FA after the staged boot-verify); direct = legacy immediate publish (emergency fallback only)"
        required: false
-        default: "auto"
+        default: "staged"
        type: choice
        options:
-          - auto
          - staged
          - direct
  workflow_call:
@@ -63,15 +62,11 @@ jobs:
    # mid-"Creating an optimized production build" while v3.8.48 had still fit in 16min.
    # This job never runs on `pull_request`, so the fork-safety clause is always true here;
    # it is kept verbatim so the expression stays greppable against ci.yml.
-    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && (github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name == github.repository)) && fromJSON('["self-hosted","omni-build"]') || 'ubuntu-latest' }}
-    outputs:
-      version: ${{ steps.resolve.outputs.version }}
-      tag: ${{ steps.resolve.outputs.tag }}
-      skip: ${{ steps.resolve.outputs.skip }}
+    runs-on: ${{ (vars.USE_VPS_RUNNER == 'true' && (github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name == github.repository)) && fromJSON('["self-hosted","omni-release"]') || 'ubuntu-latest' }}
    permissions:
      actions: read # find + download the CI run's next-build artifact for this SHA
      contents: write # gh release upload (attach SBOM to the GitHub Release)
-      id-token: write # npm provenance (GitHub Packages step)
+      id-token: write # npm provenance
      packages: write # publish to npm.pkg.github.com
    steps:
      - name: Checkout
@@ -205,11 +200,8 @@ jobs:
            exit 0
          fi
          RUN=""
-          # $RUNNER_TEMP, never /tmp: on the .113 pool /tmp is a 12 GB tmpfs (RAM). Parking
-          # this 1.3 GB artefact there took 27–32 min of the 76-min publish job — the
-          # same bytes upload from disk in 2 min. RUNNER_TEMP is per-runner and on disk.
          for candidate in $CANDIDATES; do
-            if gh run download "$candidate" --repo "$REPO" --name next-build --dir "$RUNNER_TEMP/next-build" 2>/dev/null; then
+            if gh run download "$candidate" --repo "$REPO" --name next-build --dir /tmp/next-build 2>/dev/null; then
              RUN="$candidate"
              break
            fi
@@ -219,8 +211,8 @@ jobs:
            echo "::notice::none of the candidate runs still carries next-build (1-day retention) — falling back to a full build"
            exit 0
          fi
-          tar -xzf "$RUNNER_TEMP/next-build/e2e-build.tar.gz" -C .
-          rm -rf "$RUNNER_TEMP/next-build"
+          tar -xzf /tmp/next-build/e2e-build.tar.gz -C .
+          rm -rf /tmp/next-build
          if [ -f .build/next/standalone/server.js ]; then
            echo "✅ standalone tree restored from CI run $RUN — build:cli will skip next build"
          else
@@ -273,20 +265,11 @@ jobs:
          if-no-files-found: error

      - name: Attach SBOM to GitHub Release
-        # Not only on the `release` event: the v3.8.50 package shipped through a
-        # workflow_dispatch (staged publish, 11 attempts) and this step was skipped, so the
-        # GitHub Release carried no SBOM until it was attached by hand from the run's
-        # `sbom-npm` artifact. Attach whenever a release for the published tag exists.
-        if: steps.resolve.outputs.skip != 'true' && (github.event_name == 'release' || github.event_name == 'workflow_dispatch')
+        if: steps.resolve.outputs.skip != 'true' && github.event_name == 'release'
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          TAG: ${{ github.event_name == 'release' && github.ref_name || format('v{0}', inputs.version) }}
-        run: |
-          if ! gh release view "$TAG" --repo "$GITHUB_REPOSITORY" >/dev/null 2>&1; then
-            echo "::notice::no GitHub Release for $TAG yet — SBOM stays on the sbom-npm workflow artifact"
-            exit 0
-          fi
-          gh release upload "$TAG" sbom-npm.cdx.json --repo "$GITHUB_REPOSITORY" --clobber
+          TAG: ${{ github.ref_name }}
+        run: gh release upload "$TAG" sbom-npm.cdx.json --clobber

      # WS1.2/WS1.3 (#7065 class): the artifact that is about to be published must
      # BOOT. build:cli already assembled dist/ above; this packs+installs+boots the
@@ -304,12 +287,7 @@ jobs:
      # a staged package that is never approved simply expires, with no `npm deprecate` needed.
      - name: Prove clean-install AND upgrade-over-previous both boot
        if: steps.resolve.outputs.skip != 'true'
-        # 60, not 30. This gate was added in #8953 and the 2026-08-27 v3.8.50 publish
-        # was the FIRST run to ever reach it — every earlier attempt died upstream, so
-        # its budget had never been measured against a real run. It then blew the limit
-        # on its debut: `npm pack` alone took 24m37s, leaving 5 minutes for two installs
-        # and two boots. 30 was a guess; 60 is sized to the one measurement we have.
-        timeout-minutes: 60
+        timeout-minutes: 30
        run: npm run check:install-upgrade

      # WS1.3 (D2, v3.8.49 plan): STAGED publishing by default — `npm stage publish`
@@ -331,133 +309,17 @@ jobs:
          fi
          npm --version

-      # The registry upload itself moved to the `stage-npm` job below: npm REFUSES
-      # `--provenance` from a self-hosted runner (422 "Unsupported GitHub Actions
-      # runner environment"), and the heavy verification above cannot move to a
-      # hosted one (16 GB is not enough for build:cli's next-build fallback — see
-      # this job's runs-on comment). So this job proves the bytes and hands them
-      # over; a tiny hosted job does the upload.
-      - name: Pack the verified tarball for the upload job
-        if: steps.resolve.outputs.skip != 'true'
-        env:
-          VERSION: ${{ steps.resolve.outputs.version }}
-        run: |
-          set -euo pipefail
-          # --ignore-scripts: prepublishOnly would re-run build:cli-api && build:cli,
-          # rebuilding bytes this job has already built, validated and boot-smoked.
-          npm pack --ignore-scripts
-          TARBALL="omniroute-${VERSION}.tgz"
-          test -f "$TARBALL" || { echo "expected $TARBALL to exist after npm pack" >&2; ls -la ./*.tgz || true; exit 1; }
-          echo "packed $TARBALL ($(du -h "$TARBALL" | cut -f1))"
-
-      - name: Hand the tarball to the hosted publish job
-        if: steps.resolve.outputs.skip != 'true'
-        uses: actions/upload-artifact@v7
-        with:
-          name: npm-tarball
-          path: omniroute-${{ steps.resolve.outputs.version }}.tgz
-          retention-days: 1
-          if-no-files-found: error
-
-      - name: Publish to GitHub Packages
-        if: steps.resolve.outputs.skip != 'true'
+      - name: Publish to npm (staged — owner approves with 2FA)
+        if: steps.resolve.outputs.skip != 'true' && (github.event_name != 'workflow_dispatch' || inputs.publish_mode != 'direct')
        env:
          VERSION: ${{ steps.resolve.outputs.version }}
          TAG: ${{ steps.resolve.outputs.tag }}
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        run: |
-          set -euo pipefail
-          echo "Configuring for GitHub Packages..."
-          echo "//npm.pkg.github.com/:_authToken=${GITHUB_TOKEN}" > .npmrc
-          npm pkg set name="@diegosouzapw/omniroute"
-          npm publish --registry=https://npm.pkg.github.com --tag "$TAG" \
-            || echo "⚠️ omniroute@${VERSION} might already be published on GitHub Packages."
-          echo "✅ Action finished for GitHub Packages"
-
-  # npm REFUSES `--provenance` from a self-hosted runner:
-  #   422 Unprocessable Entity - Error verifying sigstore provenance bundle:
-  #   Unsupported GitHub Actions runner environment: "self-hosted".
-  #   Only "github-hosted" runners are supported when publishing with provenance.
-  # v3.8.49 published fine because it predates USE_VPS_RUNNER being turned on
-  # (2026-08-02); v3.8.50 was the first release after it, so this had been latent
-  # for four weeks. Dropping --provenance was not an option: 3.8.49 carries a
-  # SLSA attestation and 3.8.50 must not regress that.
-  # The `publish` job cannot simply move to a hosted runner either — 16 GB is not
-  # enough for build:cli's next-build fallback. So it keeps proving the bytes and
-  # this job, which needs no memory at all, performs the upload.
-  stage-npm:
-    needs: publish
-    if: needs.publish.outputs.skip != 'true'
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-      id-token: write # npm provenance — the whole reason this job is separate
-    steps:
-      - name: Download the tarball the publish job proved
-        uses: actions/download-artifact@v8
-        with:
-          name: npm-tarball
-          path: .
-
-      - name: Setup Node.js
-        uses: actions/setup-node@v7
-        with:
-          node-version: ${{ env.NPM_PUBLISH_NODE_VERSION }}
-          registry-url: https://registry.npmjs.org
-
-      - name: Ensure npm supports staged publishing
-        run: |
-          set -euo pipefail
-          CUR=$(npm --version)
-          if ! node -e "const [a,b]='$(npm --version)'.split('.').map(Number); process.exit(a>11||(a===11&&b>=15)?0:1)"; then
-            # Pinned exact version (supply-chain: never float @latest in a publish
-            # job); bump deliberately when a newer npm is required.
-            echo "npm $CUR < 11.15 — installing pinned npm 11.15.0 for staged publishing"
-            npm install -g --ignore-scripts npm@11.15.0
-          fi
-          npm --version
-
-      # Trusted Publishing (OIDC): npm mints a short-lived credential for THIS run from
-      # GitHub's id-token — no NPM_TOKEN secret, no 2FA prompt, provenance included, and
-      # it is the bypass npm sanctions now that tokens which skip 2FA are being retired
-      # (gh.io/npm-gat-bypass2fa-deprecation). Requires the package's Trusted Publisher to
-      # be configured on npmjs.com (owner: diegosouzapw/OmniRoute, workflow
-      # npm-publish.yml) and a github-hosted runner — which is why this job exists.
-      # Without that configuration `npm publish` fails with ENEEDAUTH: re-dispatch with
-      # publish_mode=staged or direct. Automatic publishing was the flow up to v3.8.48;
-      # v3.8.49 moved to staged (WS1.3) to keep a leaked token from publishing alone —
-      # OIDC gives the same guarantee without the manual approve.
-      - name: Publish to npm (Trusted Publishing / OIDC — automatic)
-        if: github.event_name != 'workflow_dispatch' || inputs.publish_mode == 'auto'
-        env:
-          VERSION: ${{ needs.publish.outputs.version }}
-          TAG: ${{ needs.publish.outputs.tag }}
-        run: |
-          set -euo pipefail
-          TARBALL="omniroute-${VERSION}.tgz"
-          test -f "$TARBALL" || { echo "tarball $TARBALL did not arrive from the publish job" >&2; ls -la; exit 1; }
-          # Deliberately NO NODE_AUTH_TOKEN in this step: npm >= 11.5 detects the GitHub
-          # OIDC token itself. Always pass --tag explicitly (defense in depth: an older
-          # VERSION can never claim `@latest`).
-          npm publish "$TARBALL" --provenance --access public --tag "$TAG" --ignore-scripts
-          echo "✅ Published omniroute@$VERSION (dist-tag=$TAG) via Trusted Publishing"
-
-      - name: Publish to npm (staged — owner approves with 2FA)
-        # Only on an explicit request now: Trusted Publishing below is the default.
-        if: github.event_name == 'workflow_dispatch' && inputs.publish_mode == 'staged'
-        env:
-          VERSION: ${{ needs.publish.outputs.version }}
-          TAG: ${{ needs.publish.outputs.tag }}
          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}
        run: |
          set -euo pipefail
-          TARBALL="omniroute-${VERSION}.tgz"
-          test -f "$TARBALL" || { echo "tarball $TARBALL did not arrive from the publish job" >&2; ls -la; exit 1; }
          # Always pass --tag explicitly. Defense in depth: even if VERSION is
          # accidentally an older release, the historic tag will NOT claim `@latest`.
-          # --ignore-scripts: publishing a built tarball must never re-run
-          # prepublishOnly (build:cli-api && build:cli) on this small runner.
-          npm stage publish "$TARBALL" --provenance --access public --tag "$TAG" --ignore-scripts
+          npm stage publish --provenance --access public --tag "$TAG"
          {
            echo "## 📦 omniroute@$VERSION STAGED (not yet installable)"
            echo ""
@@ -473,18 +335,31 @@ jobs:
          echo "✅ Staged omniroute@$VERSION (dist-tag=$TAG) — awaiting owner 'npm stage approve'"

      - name: Publish to npm (DIRECT — emergency fallback)
-        if: github.event_name == 'workflow_dispatch' && inputs.publish_mode == 'direct'
+        if: steps.resolve.outputs.skip != 'true' && github.event_name == 'workflow_dispatch' && inputs.publish_mode == 'direct'
        env:
-          VERSION: ${{ needs.publish.outputs.version }}
-          TAG: ${{ needs.publish.outputs.tag }}
+          VERSION: ${{ steps.resolve.outputs.version }}
+          TAG: ${{ steps.resolve.outputs.tag }}
          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}
        run: |
          set -euo pipefail
-          TARBALL="omniroute-${VERSION}.tgz"
-          test -f "$TARBALL" || { echo "tarball $TARBALL did not arrive from the publish job" >&2; ls -la; exit 1; }
-          npm publish "$TARBALL" --provenance --access public --tag "$TAG" --ignore-scripts
+          npm publish --provenance --access public --tag "$TAG"
          echo "✅ Published omniroute@$VERSION (dist-tag=$TAG) [DIRECT mode]"

+      - name: Publish to GitHub Packages
+        if: steps.resolve.outputs.skip != 'true'
+        env:
+          VERSION: ${{ steps.resolve.outputs.version }}
+          TAG: ${{ steps.resolve.outputs.tag }}
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -euo pipefail
+          echo "Configuring for GitHub Packages..."
+          echo "//npm.pkg.github.com/:_authToken=${GITHUB_TOKEN}" > .npmrc
+          npm pkg set name="@diegosouzapw/omniroute"
+          npm publish --registry=https://npm.pkg.github.com --tag "$TAG" \
+            || echo "⚠️ omniroute@${VERSION} might already be published on GitHub Packages."
+          echo "✅ Action finished for GitHub Packages"
+
  publish-opencode-plugin:
    runs-on: ubuntu-latest
    permissions:
--- a/.github/workflows/scorecard.yml
+++ b/.github/workflows/scorecard.yml
@@ -4,15 +4,12 @@ on:
  schedule:
    - cron: "27 7 * * 1"
  push:
-    # Scorecard only accepts the DEFAULT branch — here the active release/vX.Y.Z,
-    # not `main`. The job below guards on it so a push to any other branch skips.
-    branches: ["main", "release/**"]
+    branches: ["main"]

 permissions: read-all

 jobs:
  analysis:
-    if: ${{ github.event_name != 'push' || github.ref == format('refs/heads/{0}', github.event.repository.default_branch) }}
    name: Scorecard analysis
    runs-on: ubuntu-latest
    permissions:
--- a/.gitignore
+++ b/.gitignore
@@ -293,6 +293,3 @@ docker-compose.yml.bak
 # Ad-hoc test sandboxes (never tracked — may contain local DBs)
 /.sandbox/
 .aider*
-
-# check:install-upgrade work trees (~12 GB, disposable)
-/.install-upgrade/
--- a/.mailmap
+++ b/.mailmap
@@ -1,40 +0,0 @@
-# .mailmap — canonical author identities for git log/shortlog/blame.
-#
-# Why this file exists: between 2026-08-13 and 2026-08-26 this checkout carried a
-# `git config --local` whose user.name was one contributor's ("Xiangzhe" / @xz-dev)
-# and whose user.email was ANOTHER contributor's (@backryun). Every commit produced
-# on this machine in that window was therefore signed with @backryun's address —
-# 237 commits, all in the -0300 timezone, while @backryun's own work commits from
-# +0900 and continued normally throughout. The local override was removed on
-# 2026-08-26; this file repairs the RECORD without rewriting published history
-# (those commits live on release/v3.8.50 and release/v3.8.51, which other sessions
-# and open PRs build on — a rewrite would force-push both and orphan the v3.8.50 tag).
-#
-# Format: Canonical Name <canonical@email> Commit Name <commit@email>
-
-# --- Maintainer: several addresses used over the project's life ---
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> <diegosouza.pw@gmail.com>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> <diegosouza.pw@outlook.com>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> <diegosouzapw@users.noreply.github.com>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> Diego Rodrigues de Sa e Souza <8016841+diegosouzapw@users.noreply.github.com>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> Diego Souza <8016841+diegosouzapw@users.noreply.github.com>
-
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> <diego.souza.pw@gmail.com>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> <souzamiriamrodrigues790@gmail.com>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> <diego.souza@cdwasolutions.com.br>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> <diegosouzapw@devbox.local>
-
-# --- The misattribution window: name Xiangzhe + @backryun's email, from -0300.
-# These are maintainer/session commits, NOT @backryun's contributions.
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> Xiangzhe <bakryun0718@proton.me>
-diegosouzapw <8016841+diegosouzapw@users.noreply.github.com> Xiangzhe <diegosouza.pw@gmail.com>
-
-# --- Xiangzhe (@xz-dev) — a distinct contributor; keep their own work intact ---
-Xiangzhe <32761048+xz-dev@users.noreply.github.com> <xz-dev@users.noreply.github.com>
-Xiangzhe <32761048+xz-dev@users.noreply.github.com> <xiangzhedev@gmail.com>
-
-# --- @backryun's own alternate addresses (their real work, kept intact) ---
-backryun <24198422+backryun@users.noreply.github.com> <bakryun0718@proton.me>
-backryun <24198422+backryun@users.noreply.github.com> <backryun@daonlab.local>
-backryun <24198422+backryun@users.noreply.github.com> <busan011@ormbiz.co.kr>
-backryun <24198422+backryun@users.noreply.github.com> <backryun@users.noreply.github.com>
--- a/.trivyignore
+++ b/.trivyignore
@@ -19,12 +19,4 @@
 # Keep this list SHORT and reviewed every release. Prefer fixing (rebuild on a
 # patched base / bump the dep) over suppressing. Stale entries are debt.
 #
-# CVE-2025-68121 — Go stdlib crypto/tls (session-resumption certificate validation)
-#   inside the PREBUILT bogdanfinn/tls-client v1.15.1 .so that tls-client-node's
-#   postinstall downloads (built with go 1.24.1; fixed in 1.24.13). No upstream
-#   rebuild exists (v1.15.1 is still the latest release) and nothing in this repo
-#   can bump it. The binary is only loaded by the browser-TLS web-provider
-#   executors (claude-web / grok-web / lmarena / perplexity-web / notion-web),
-#   whose handshakes go through utls. Tracking issue: #12084. Revisit at the next
-#   tls-client release or base-image bump and BEFORE the v3.8.51 tag (2026-09-15).
-CVE-2025-68121
+# (No accepted-risk suppressions at present — ignore-unfixed covers the noise.)
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -56,7 +56,7 @@ Repository map and Reference Documentation sections below.
 | 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 (160 migrations)                                                                                                                                    |
+| Database      | `src/lib/db/`           | SQLite domain modules (159 migrations)                                                                                                                                    |
 | Domain/Policy | `src/domain/`           | Policy engine, cost rules, fallback logic                                                                                                                                 |
 | MCP Server    | `open-sse/mcp-server/`  | 110 tools (44 canonical + memory/skill/GitHub/pool/gamification/plugin/Notion/Obsidian/local-corpus/RTK modules), 3 transports (stdio / SSE / Streamable HTTP), 33 scopes |
 | A2A Server    | `src/lib/a2a/`          | JSON-RPC 2.0 agent protocol                                                                                                                                               |
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -93,67 +93,6 @@

 _Living section — regenerated 2026-08-12 from all cycle commits (cycle open `ed2db6cb19` → tip). Bullets carry the merged PR and its author; direct pushes listed separately._

-### 📊 Release by the numbers
-
-| | |
-| --- | ---: |
-| 👥 People who contributed | **248** |
-| 📝 Commits in the cycle | **1,714** |
-| 🔀 Pull requests referenced | **1,666** |
-| 📋 Changelog entries | **1,182** |
-| 🙌 Contributors credited in entries | **256** |
-| 🤖 Automated dependency commits | 22 |
-
-**Entries by type**
-
-| Type | Count |
-| --- | ---: |
-| 🐛 Fixes | 779 |
-| ✨ Features | 169 |
-| 📚 Docs | 29 |
-| 🧹 Chore | 27 |
-| 🧪 Tests | 15 |
-| ♻️ Refactor | 5 |
-| ⚡ Performance | 3 |
-| providers | 2 |
-| 🔒 Security | 2 |
-| ⚙️ CI | 2 |
-| deps | 2 |
-| maint | 2 |
-
-### 🏆 Top 25 contributors this cycle
-
-_By commits in `ed2db6cb19..v3.8.50`, author identities consolidated via `.mailmap`. Bots excluded._
-
-| # | Contributor | Commits |
-| ---: | --- | ---: |
-| 🥇 | diegosouzapw | 738 |
-| 🥈 | backryun | 88 |
-| 🥉 | Dizzle | 66 |
-| 4 | Ravi Tharuma | 52 |
-| 5 | Markus Hartung | 48 |
-| 6 | Bob.Hou | 42 |
-| 7 | Rouzbeh† | 38 |
-| 8 | Xiangzhe | 31 |
-| 9 | Paco Cartones | 28 |
-| 10 | Nguyen Thanh Dat | 23 |
-| 11 | Aman | 22 |
-| 12 | Will Gordon | 19 |
-| 13 | 小妍儿 ✨ | 17 |
-| 14 | adevwithpurpose | 16 |
-| 15 | Andrew B. | 10 |
-| 16 | NOXX - Commiter | 10 |
-| 17 | ignamiranda | 10 |
-| 18 | Jonathan Bailey | 9 |
-| 19 | Ke Jin | 9 |
-| 20 | Austin Liu | 8 |
-| 21 | Chewji | 8 |
-| 22 | Prudhvi Vuda | 7 |
-| 23 | benzntech | 7 |
-| 24 | rinseaid | 7 |
-| 25 | stanley | 7 |
-
-
 ### ✨ New Features
 - **feat(search):** first-class X Search provider (`x-search`) on `POST /v1/search` and MCP `omniroute_x_search` using SuperGrok / xAI server-side `x_search`. Explicit provider or `search_type: "x"` only — never auto-selected for web. Reuses `xai-oauth` / `xao` / `xai` credentials. Not the X Developer Platform MCP. ([#10985](https://github.com/diegosouzapw/OmniRoute/issues/10985))
 - **feat(core):** add Layer A capability filter at router (#5696)
--- a/36
+++ b/36
@@ -184,29 +184,19 @@ ENV NODE_OPTIONS="--max-old-space-size=${OMNIROUTE_BUILD_MEMORY_MB}"
 # silently leaving no standalone bundle. Next derives the worker count from
 # CIRCLE_NODE_TOTAL (workers = N-1). (#10060)
 #
-# Lowered 8 → 3 (7 workers → 2) in #11419, then 3 → 2 (2 workers → 1) in #7518.
-# Every page-data worker inherits NODE_OPTIONS above, so the ceiling is per
-# PROCESS, not per build: 7 workers on a 16 GB GitHub runner (ubuntu-24.04 /
-# ubuntu-24.04-arm, 4 vCPU) exhausted the host and buildkit failed the whole
-# step with `ResourceExhausted: ... cannot allocate memory`. The compile phase
-# always finished ("✓ Compiled successfully in 4.2min"); the kernel killed the
-# build right after "Collecting page data using N workers".
-#
-# #11419's first fix (8 → 3) modeled the per-worker peak as an INFERENCE
-# (2560 MB, guessed from "7 workers didn't fit") and assumed the parent
-# process's RSS tracked the V8 heap ceiling. Both assumptions were wrong: a
-# live VPS reproduction (issue #7518, dmesg OOM-killer report) measured the
-# real per-process RSS directly at ~4.5 GB, independent of the NODE_OPTIONS
-# heap flag (Turbopack itself is native/Rust, outside the V8 heap) — and it
-# applies to the parent process too, not just workers. 2 workers (3 processes
-# × 4.5 GB = 13.5 GB) still didn't fit the 12.288 GB (75%) budget on a 16 GB
-# runner, matching the still-live publish failures after #11419 merged. 1
-# worker (2 processes × 4.5 GB = 9 GB) fits with headroom to spare.
-# tests/unit/docker-build-memory-budget.test.ts does the arithmetic against
-# the measured figure and fails if either knob is raised past what a 16 GB
-# runner holds. Override for a big builder: `--build-arg
-# OMNIROUTE_BUILD_WORKERS=8`.
-ARG OMNIROUTE_BUILD_WORKERS=2
+# Lowered 8 → 3 (7 workers → 2). Every page-data worker inherits NODE_OPTIONS
+# above, so the ceiling is per PROCESS, not per build: 7 workers on a 16 GB
+# GitHub runner (ubuntu-24.04 / ubuntu-24.04-arm, 4 vCPU) exhausted the host and
+# buildkit failed the whole step with `ResourceExhausted: ... cannot allocate
+# memory`. The compile phase always finished ("✓ Compiled successfully in
+# 4.2min"); the kernel killed the build right after "Collecting page data using
+# 7 workers". It was intermittent for a while and went 100% on 2026-08-22, which
+# is what a threshold being crossed by ordinary codebase growth looks like.
+# tests/unit/docker-build-memory-budget.test.ts does the arithmetic and fails if
+# either knob is raised past what a 16 GB runner holds. 2 workers also stops
+# oversubscribing the runner's 4 vCPU, which 7 did. Override for a big builder:
+# `--build-arg OMNIROUTE_BUILD_WORKERS=8`.
+ARG OMNIROUTE_BUILD_WORKERS=3
 ENV CIRCLE_NODE_TOTAL=${OMNIROUTE_BUILD_WORKERS}

 COPY . ./
--- a/Dockerfile.bun
+++ b/Dockerfile.bun
@@ -1,5 +1,5 @@
 # ── Multi-stage Dockerfile for Native Bun Runtime (web-latest-bun) ───────────
-FROM oven/bun:1.4.0-slim AS base
+FROM oven/bun:1.3.14-slim AS base
 WORKDIR /app

 RUN apt-get update \
@@ -34,10 +34,8 @@ RUN if [ -f "node_modules/tls-client-node/scripts/postinstall.js" ]; then \
      bun node_modules/tls-client-node/scripts/postinstall.js || true; \
    fi

-# Turbopack is supported on Bun 1.4+ (Next 16.3); override via
-# --build-arg OMNIROUTE_USE_TURBOPACK=0 to force the webpack fallback.
-ARG OMNIROUTE_USE_TURBOPACK=1
-ENV OMNIROUTE_USE_TURBOPACK=${OMNIROUTE_USE_TURBOPACK}
+# Disable Turbopack for Bun builder stage (Turbopack V8 internal worker bindings require Node)
+ENV OMNIROUTE_USE_TURBOPACK=0

 ARG OMNIROUTE_BASE_PATH=""
 ENV OMNIROUTE_BASE_PATH=$OMNIROUTE_BASE_PATH
@@ -48,33 +46,6 @@ ENV DASHBOARD_ALLOW_EMBED=$DASHBOARD_ALLOW_EMBED
 ENV NEXT_TELEMETRY_DISABLED=1
 ENV NODE_ENV=production

-# Cap the Next.js build heap and page-data worker pool inside the Bun image the
-# same way the node Dockerfile does (#10060/#11419/#7518). Without these knobs
-# Next falls back to its defaults: worker pool = os.cpus()-1 (3 on the 4-vCPU
-# GitHub runner) and an 8 GB V8 heap ceiling per process. 4+ V8 processes at
-# multi-GB each blow past the 16 GB runner, the cgroup OOM killer SIGKILLs a
-# build worker mid-compile, and buildx fails the step with `ResourceExhausted:
-# ... cannot allocate memory` — every Bun image published on main since the -bun
-# targets landed (#11709, #11039).
-#
-# The per-process peak is a MEASURED ~4.5 GB RSS (dmesg OOM-killer report,
-# #7518), independent of NODE_OPTIONS — Turbopack is native/Rust and compiles
-# outside the V8 heap — and it applies to the parent process too, so 2 page-data
-# workers (3 processes × 4.5 GB ≈ 13.5 GB) do not fit the 12.288 GB (75%)
-# budget either. Both images therefore default to OMNIROUTE_BUILD_WORKERS=2
-# (1 page-data worker): 2 processes × 4.5 GB ≈ 9 GB fits with headroom (#11663).
-# The default Turbopack path keeps the compile outside the V8 heap, but the
-# guards must hold for the webpack fallback (OMNIROUTE_USE_TURBOPACK=0) too, so
-# they are wired exactly like the node image.
-#
-# NODE_OPTIONS propagates to the spawned `next build` child and its workers
-# (build-next-isolated.mjs → resolveNextBuildEnv spreads process.env), so the
-# ceiling is per PROCESS, not per build.
-ARG OMNIROUTE_BUILD_MEMORY_MB=6144
-ENV NODE_OPTIONS="--max-old-space-size=${OMNIROUTE_BUILD_MEMORY_MB}"
-ARG OMNIROUTE_BUILD_WORKERS=2
-ENV CIRCLE_NODE_TOTAL=${OMNIROUTE_BUILD_WORKERS}
-
 # Bun native Next.js build execution
 RUN bun run --quiet build

--- a/README.md
+++ b/README.md
@@ -1202,7 +1202,7 @@ Métricas canônicas em 2026-08-24: **1.029 vídeos únicos** · **11.132.922 vi
  <tr><td nowrap><b>Runtime</b></td><td>Node.js 22.x / 24.x LTS — <code>&gt;=22.22.2 &lt;23 || &gt;=24.0.0 &lt;27</code></td></tr>
  <tr><td nowrap><b>Language</b></td><td>TypeScript 6.0 — <b>100% TypeScript</b> across <code>src/</code> and <code>open-sse/</code> (zero <code>any</code> in core since v2.0)</td></tr>
  <tr><td nowrap><b>Framework</b></td><td>Next.js 16 + React 19 + Tailwind CSS 4</td></tr>
-  <tr><td nowrap><b>Database</b></td><td>better-sqlite3 (SQLite, WAL journaling) + LowDB (JSON legacy) — 120 domain modules, 160 migrations</td></tr>
+  <tr><td nowrap><b>Database</b></td><td>better-sqlite3 (SQLite, WAL journaling) + LowDB (JSON legacy) — 120 domain modules, 159 migrations</td></tr>
  <tr><td nowrap><b>Memory</b></td><td>SQLite FTS5 full-text + int8-quantized vector embeddings, typed decay</td></tr>
  <tr><td nowrap><b>Schemas</b></td><td>Zod 4 — MCP tool I/O validation + API contracts</td></tr>
  <tr><td nowrap><b>Protocols</b></td><td>MCP (stdio / HTTP / SSE) + A2A v0.3 (JSON-RPC 2.0 + SSE)</td></tr>
--- a/changelog.d/features/npm-trusted-publishing-oidc.md
+++ b/changelog.d/features/npm-trusted-publishing-oidc.md
@@ -1,4 +0,0 @@
- The npm publish is automatic again, through npm Trusted Publishing (OIDC): the hosted
-  `stage-npm` job publishes with a short-lived credential minted from GitHub's id-token —
-  no `NPM_TOKEN`, no 2FA prompt, provenance attached. `publish_mode=staged` (owner
-  approves with 2FA) and `direct` (token) remain available on `workflow_dispatch`.
--- a/changelog.d/fixes/11719-bun-image-turbopack-memory-guards.md
+++ b/changelog.d/fixes/11719-bun-image-turbopack-memory-guards.md
@@ -1 +0,0 @@
- **fix(docker):** bump the Bun image to 1.4.0, enable Turbopack on Bun, and port the node image's build memory guards so the `-bun` container builds fit the 16 GB GitHub runner instead of dying with `cannot allocate memory` ([#11719](https://github.com/diegosouzapw/OmniRoute/pull/11719)). Both images now default `OMNIROUTE_BUILD_WORKERS` to `2` (1 page-data worker) against the measured ~4.5 GB per-process RSS budget (#7518/#11663).
--- a/changelog.d/fixes/11741-install-upgrade-schema-convergence.md
+++ b/changelog.d/fixes/11741-install-upgrade-schema-convergence.md
@@ -1,8 +0,0 @@
- **fix(db):** `model_capabilities` is created by a migration instead of lazily on the first
-  models.dev sync, so a clean install and an upgraded install converge on the same schema
-  regardless of which features have run
- **fix(ci):** `check:install-upgrade` now fails on an `npm` install truncated by ENOSPC
-  (npm reports it as a warning and still exits 0), authenticates its health probe so the
-  version assertion works against the hardened health payload, frees the clean-install tree
-  before the upgrade phase, and no longer reports a schema divergence computed from a boot
-  that never served
--- a/changelog.d/fixes/11856-npm-payload-nft-manifests.md
+++ b/changelog.d/fixes/11856-npm-payload-nft-manifests.md
@@ -1,4 +0,0 @@
- Excluded Next.js Node File Trace manifests (`*.nft.json`) from the published npm
-  tarball. They are build-time metadata and are never read while serving, but had
-  grown to 668.7 MB — 61% of the package — which pushed the upload past the
-  registry limit and made `npm publish` fail with `413 Payload Too Large`.
--- a/changelog.d/fixes/11866-alibaba-allowlist-test-timebomb.md
+++ b/changelog.d/fixes/11866-alibaba-allowlist-test-timebomb.md
@@ -1,5 +0,0 @@
- Fixed the Alibaba free-tier allowlist test that went red on its own once the
-  shipped catalog's `validUntil` (2026-08-27) passed, leaving every PR and `main`
-  with a failing `Unit Tests (1/8)`. The test now builds its own packs with dates
-  it controls, and covers the expired-pack fallback that production has actually
-  been serving.
--- a/changelog.d/fixes/11868-npm-provenance-hosted-runner.md
+++ b/changelog.d/fixes/11868-npm-provenance-hosted-runner.md
@@ -1,4 +0,0 @@
- Split the npm registry upload into its own GitHub-hosted job. npm refuses
-  `--provenance` from a self-hosted runner (`422 ... Only "github-hosted" runners
-  are supported`), which blocked the v3.8.50 publish; the heavy verification
-  cannot move to a hosted runner, so it now hands the proven tarball over instead.
--- a/changelog.d/fixes/v3850-electron-lockfile-build-ref.md
+++ b/changelog.d/fixes/v3850-electron-lockfile-build-ref.md
@@ -1 +0,0 @@
- Electron release: `electron/package-lock.json` regained the optional `electron-builder-squirrel-windows` subtree (13 entries) that `npm ci` had been refusing as out of sync — the Linux desktop leg died on it — and `electron-release.yml` gained a `build_ref` dispatch input so a release whose tag was cut with the broken lock can have its assets rebuilt from the repaired line
--- a/changelog.d/fixes/v3850-electron-release-assets.md
+++ b/changelog.d/fixes/v3850-electron-release-assets.md
@@ -1 +0,0 @@
- Electron release workflow: the `publish-npm` job now grants `actions: read` to the reusable `npm-publish.yml` it calls (its `publish` job requests it), which is what made GitHub refuse the whole v3.8.50 run at startup and ship the release with zero desktop assets; a `workflow_dispatch` now builds the requested tag instead of the dispatching branch and can skip the npm leg (`publish_npm=false`) when only re-attaching assets
--- a/changelog.d/fixes/v3850-sbom-attach-on-dispatch.md
+++ b/changelog.d/fixes/v3850-sbom-attach-on-dispatch.md
@@ -1 +0,0 @@
- npm publish workflow: the CycloneDX SBOM is attached to the GitHub Release on `workflow_dispatch` publishes too (when a release for the tag exists), not only on the `release` event — v3.8.50 shipped through a staged dispatch and its release carried no SBOM until it was attached by hand from the run's `sbom-npm` artifact
--- a/changelog.d/maintenance/11866-config-expiry-time-bomb-test.md
+++ b/changelog.d/maintenance/11866-config-expiry-time-bomb-test.md
@@ -1,4 +0,0 @@
- Added a unit test that fails seven days before any dated pack under `config/`
-  (`validUntil` and sibling keys) lapses, naming the file and key. The Alibaba
-  free-tier pack expired on 2026-08-27 and turned every PR red the next morning
-  with no commit involved; renewal now happens on someone's terms, not the clock's.
--- a/changelog.d/maintenance/11878-check-workflows-provenance-self-hosted.md
+++ b/changelog.d/maintenance/11878-check-workflows-provenance-self-hosted.md
@@ -1,3 +0,0 @@
- `check:workflows` now fails (under `--strict`/`--ratchet`) when any job routed to a
-  self-hosted runner publishes with `--provenance` — npm rejects that with `422` at the
-  registry, which in v3.8.50 only surfaced after the tag and Docker images were public.
--- a/changelog.d/maintenance/11892-runner-janitor-act-not-advise.md
+++ b/changelog.d/maintenance/11892-runner-janitor-act-not-advise.md
@@ -1,5 +0,0 @@
- `scripts/ops/runner-janitor.sh` now proves a path is idle with one `lsof`
-  snapshot and removes stale leftovers itself (tmpfs after 3 h — it is RAM — disk
-  after 24 h), kills orphan `next-build` processes, prunes checkouts of stopped
-  runners, and alerts on memory pressure; `--dry-run` shows exactly what it would
-  do. `docs/ops/RUNNER_BOX.md` reconciled to the measured box (31 GB, 10 listeners).
--- a/changelog.d/maintenance/11896-ci-artifact-download-to-disk.md
+++ b/changelog.d/maintenance/11896-ci-artifact-download-to-disk.md
@@ -1,5 +0,0 @@
- The `next-build` artefact (1.3 GB) is now written and read under `$RUNNER_TEMP`
-  (per-runner, on disk) instead of `/tmp`, which on the self-hosted pool is a
-  12 GB tmpfs in RAM. Landing it there took 27–32 of the publish job's 76 minutes,
-  and the fixed `/tmp/e2e-build.tar.gz` name let E2E jobs on different runners
-  overwrite each other's download.
--- a/changelog.d/maintenance/11897-ci-heavy-build-lane.md
+++ b/changelog.d/maintenance/11897-ci-heavy-build-lane.md
@@ -1,3 +0,0 @@
- The CI `build` job now runs in two concurrency lanes — `main` and pull requests —
-  so a release build is never queued behind (or OOM-killed beside) PR builds on the
-  self-hosted pool, which holds one `next-build` comfortably and two at the edge.
--- a/changelog.d/maintenance/ci-coverage-codecov-step-timeout.md
+++ b/changelog.d/maintenance/ci-coverage-codecov-step-timeout.md
@@ -1 +0,0 @@
- `Coverage` job on `ci.yml`: the informational Codecov upload gets its own 5-minute ceiling and `continue-on-error`, and the job budget grows from 20 to 30 minutes (the 8-shard c8 merge alone takes ~10) — a stalled upload no longer ends the job `cancelled` and drags a fully green `main` run's conclusion down with it
--- a/changelog.d/maintenance/ci-omni-build-runner-label.md
+++ b/changelog.d/maintenance/ci-omni-build-runner-label.md
@@ -1,4 +0,0 @@
- Every CI job that runs a `next build` (`build`, the npm `publish`, both release-green
-  validations) now targets the `omni-build` runner label, which only two of the eight
-  self-hosted runners carry. The box holds one build comfortably and two at the edge; a
-  third now queues on GitHub instead of being OOM-killed by the kernel.
--- a/config/quality/install-upgrade-allowlist.json
+++ b/config/quality/install-upgrade-allowlist.json
@@ -1,6 +1,6 @@
 {
-  "_doc": "Tables that exist ONLY in databases upgraded from an older version. Two causes, and they call for different fixes: (a) residue whose CREATE left the migration set in some past cycle but survives where it already existed — allowlist it here; (b) a table created LAZILY at runtime with `CREATE TABLE IF NOT EXISTS` inside a feature code path — whether a database has it depends on whether that feature ran, so it diverges by TIMING and can show up on EITHER side. Fix (b) with a migration instead of an entry here (see src/lib/db/migrations/163_model_capabilities.sql); an allowlist entry only hides it in one direction. Either way, recorded so check-install-upgrade.mjs can still fail on a NEW divergence. The opposite direction (a table a clean install creates but an upgrade does not) is NEVER allowlisted: it means every existing user is missing structure the code expects.",
+  "_doc": "Tables that exist ONLY in databases upgraded from an older version — residue whose CREATE left the migration set in some past cycle but survives where it already existed. Harmless (nothing references them), but recorded here so check-install-upgrade.mjs can still fail on a NEW divergence. The opposite direction (a table a clean install creates but an upgrade does not) is NEVER allowlisted: it means every existing user is missing structure the code expects.",
  "residualTables": {
-    "cache_metrics": "Measured 2026-07-30 on a real 3.8.48 install upgraded to 3.8.49 (VPS .16, 165 MB database, 114 → 117 tables). Present in upgraded databases, absent from clean installs. Cause identified 2026-08-27: it is case (b) above — created lazily by `ensureCacheMetricsTable()` at src/lib/semanticCache.ts:34, never by a migration, so it appears only where the semantic cache has run. No code path referenced it during the upgrade (zero `no such table` in 150 log lines, both installs healthy). Left as an allowlist entry rather than promoted to a migration or dropped: unlike model_capabilities it did not block a release, and creating a table for a subsystem we cannot prove is live is not a change to make blind. Revisit when the cache subsystem is next touched."
+    "cache_metrics": "Measured 2026-07-30 on a real 3.8.48 install upgraded to 3.8.49 (VPS .16, 165 MB database, 114 → 117 tables). Present in upgraded databases, absent from clean installs. No code path referenced it during the upgrade (zero `no such table` in 150 log lines, both installs healthy). Left in place rather than dropped: a DROP migration on a table we cannot prove is unused everywhere is the riskier change. Revisit when the cache subsystem is next touched."
  }
 }
--- a/config/quality/quality-baseline.json
+++ b/config/quality/quality-baseline.json
@@ -169,7 +169,7 @@
      "dedicatedGate": true
    },
    "zizmorFindings": {
-      "value": 194,
+      "value": 192,
      "_rebaseline_2026_08_20_radar_export_workflow": "190 -> 192 (+2). Workflow novo `.github/workflows/radar-export.yml` (passo 10 do go-live do Radar: publica o export estável do catálogo como asset de release para o servidor privado baixar via RADAR_EXPORT_URL). Os +2 são unpinned-uses @vN: actions/checkout@v7 + actions/setup-node@v7 — a MESMA convenção deliberada de todos os workflows (ver _scanner_harden_workflows_2026_06_16); fixar por SHA só este violaria a convenção. O findings artipacked do checkout foi CORRIGIDO com `persist-credentials: false` (o job publica via GH_TOKEN em `gh release`, não usa a credencial do checkout). Nenhuma classe nova de template-injection / cache-poisoning / dangerous-triggers. Medido local com zizmor 1.25.2 via `node scripts/check/check-workflows.mjs --ratchet` = 191; +1 do delta conhecido do runner (ver _rebaseline_2026_07_28_ci_runner_delta: o runner enxerga 1 unpinned-uses @vN a mais que o devbox no mesmo commit; a baseline segue o runner) => 192.",
      "_rebaseline_2026_07_20_aliasresolver_hook_split_7808": "175 -> 176 (+1). Companion to PR #7808 (CodeQL js/incomplete-url-substring-sanitization fix in bin/aliasResolver.mjs). The +1 is NOT caused by this PR's code changes (bin/* is not a workflow file) — it is a pre-existing drift that surfaced because the ratchet gate runs on this PR's CI: the zizmor scanner version on the GitHub runner gained a new rule (or extended an existing one) since the v3.8.49 baseline was seeded on 2026-07-17. Breakdown: the new finding is an unpinned-uses @vN class item on one of the existing workflows (same deliberate convention as _scanner_harden_workflows_2026_06_16 — @vN is intentional, SHA-pinning only this one would violate the convention). No new template-injection/artipacked/cache-poisoning/dangerous-triggers classes introduced. Measured by the Quality Gates (Extended) job on run 29713001401 = 176, baseline was 175. Note: by the time this landed on release/v3.8.49, the baseline was already at 176 via _rebaseline_2026_07_17_combo_recovery_hints — this entry is kept as historical record; no further bump applied.",
      "_rebaseline_2026_07_17_v3849_release": "169 -> 175 (+6). Cycle workflow drift (v3.8.48/v3.8.49): npm-publish.yml (new, WS1.3 #7092), electron-release.yml, nightly-compat.yml, nightly-release-green.yml, CI restructures (#7501 full-history base fetch, #7355 main-green, #7202 merge-queue gates, Trunk/Codecov). Breakdown vs v3.8.47: +3 unpinned-uses (@vN convention, deliberate per _scanner_harden_workflows_2026_06_16), +2 cache-poisoning (artifact upload/cache in the OWN electron-release/npm-publish RELEASE workflows -- operator-controlled, not fork-PR exploitable), +1 excessive-permissions (nightly-compat.yml permissions:issues). No new template-injection/artipacked/dangerous-triggers. Measured with zizmor 1.25.2 via `node scripts/check/check-workflows.mjs --ratchet` = 175 on da3a0be69.",
@@ -180,8 +180,7 @@
      "_rebaseline_2026_06_23_v3834_release": "152 -> 155 (+3). The 3 new unpinned-uses are in .github/workflows/nightly-release-green.yml (added by #4622 this cycle): actions/checkout@v7, actions/setup-node@v6, actions/upload-artifact@v4 — the SAME deliberate @vN convention as ci.yml's own checkout@v7/setup-node@v6 and every other workflow (see _scanner_harden_workflows_2026_06_16 + _zizmor_rebaseline_2026_06_20_ci_build_artifact_reuse). SHA-pinning only this workflow would violate the convention. The workflow-lint ratchet does NOT run on PR->release fast-gates, so it surfaced only on the release PR; measured locally via `npm run check:workflows -- --ratchet` = 155. No new template-injection/artipacked/cache-poisoning.",
      "_rebaseline_2026_07_13_v3847_release_preflight": "159 -> 169 (+10). Findings from cycle-merged workflow changes: #6716 (PR gate restructure), #6781 (unit fast-path shard 2->4), #6788 (TIA tsx loader split), #6881 (electron-updater latest.yml manifests in release assets) — same deliberate @vN unpinned-uses convention as prior rebaselines; no new template-injection/artipacked/cache-poisoning classes. Measured via `npm run check:workflows -- --ratchet` = 169 on the v3.8.47 release pre-flight.",
      "_rebaseline_2026_07_28_v3849_release_preflight": "176 -> 189 (+13). Pre-flight de fechamento da v3.8.49 (934 commits no ciclo). Deriva de workflow: 1 workflow novo (build-rinseaid-image.yml) mais os bumps de action do Dependabot ao longo do ciclo — todos da MESMA classe unpinned-uses @vN, convenção deliberada do repo (ver _scanner_harden_workflows_2026_06_16); fixar por SHA só estes violaria a convenção. Nenhuma classe nova de template-injection / artipacked / cache-poisoning / dangerous-triggers. Nesta mesma passada foram CORRIGIDAS 3 diretivas shellcheck malformadas (SC1125: `# shellcheck disable=SC2086 — texto`, em que o travessão invalida o par key=value) em ci.yml e nightly-release-green.yml. Medido com zizmor 1.25.2 via `npm run check:workflows -- --ratchet` = 189.",
-      "_rebaseline_2026_07_28_ci_runner_delta": "189 -> 190 (+1). Medido 189 no devbox e 190 no runner do GitHub no MESMO commit (run 30396592013, job Quality Gates (Extended)) — mesma classe já registrada em _rebaseline_2026_07_20_aliasresolver_hook_split_7808: a versão do zizmor no runner enxerga uma finding a mais que a local, sempre da classe unpinned-uses @vN. O valor do runner é o que o gate compara, então a baseline segue o runner.",
-      "_rebaseline_2026_08_28_npm_publish_hosted_stage_job": "192 -> 194 (+2). Job novo `stage-npm` em .github/workflows/npm-publish.yml: o npm RECUSA `--provenance` vindo de runner self-hosted (422 \"Unsupported GitHub Actions runner environment\"), e o job `publish` nao pode migrar para runner hospedado porque 16 GB nao bastam para o fallback next-build do build:cli (documentado no proprio runs-on). A separacao foi a unica saida que preserva a atestacao SLSA que a 3.8.49 ja tem. Os +2 sao da MESMA convencao deliberada de todos os workflows (ver _scanner_harden_workflows_2026_06_16): unpinned-uses @vN em actions/download-artifact@v8 + actions/setup-node@v7, mais o cache-poisoning que o proprio setup-node@v7 ja gera nos outros 2 jobs deste MESMO arquivo (linhas 85 e 463) e que ja esta na baseline. Fixar por SHA so este job violaria a convencao. Nenhuma classe nova: zero template-injection / artipacked / dangerous-triggers / excessive-permissions — o job declara apenas contents:read + id-token:write, que e o minimo para a proveniencia. Medido pelo job Quality Gates (Extended) no run 33162... da PR #11877 = 194."
+      "_rebaseline_2026_07_28_ci_runner_delta": "189 -> 190 (+1). Medido 189 no devbox e 190 no runner do GitHub no MESMO commit (run 30396592013, job Quality Gates (Extended)) — mesma classe já registrada em _rebaseline_2026_07_20_aliasresolver_hook_split_7808: a versão do zizmor no runner enxerga uma finding a mais que a local, sempre da classe unpinned-uses @vN. O valor do runner é o que o gate compara, então a baseline segue o runner."
    },
    "vulnCount": {
      "value": 22,
--- a/docs/guides/DOCKER_GUIDE.md
+++ b/docs/guides/DOCKER_GUIDE.md
@@ -226,22 +226,16 @@ Three build args control what the `builder` stage costs. They are build-time onl
 | --------------------------- | ------- | ----------------------------------------------------------------------------------- |
 | `OMNIROUTE_USE_TURBOPACK`   | `1`     | `0` builds with webpack instead. Lower peak memory, slower.                         |
 | `OMNIROUTE_BUILD_MEMORY_MB` | `6144`  | V8 heap ceiling (`--max-old-space-size`) for the spawned `next build`.              |
-| `OMNIROUTE_BUILD_WORKERS`   | `2`     | Feeds `CIRCLE_NODE_TOTAL`; Next derives `workers = N - 1` for page-data collection. |
+| `OMNIROUTE_BUILD_WORKERS`   | `3`     | Feeds `CIRCLE_NODE_TOTAL`; Next derives `workers = N - 1` for page-data collection. |

 `OMNIROUTE_BUILD_WORKERS` is the one to raise on a big builder and the one to
 suspect when a constrained build dies **after** `✓ Compiled successfully`. Each
-page-data worker is its own process, and so is the parent `next build` itself;
-a live VPS reproduction (issue #7518) measured each process's peak RSS at
-~4.5 GB independent of the `NODE_OPTIONS` heap flag (Turbopack compiles in
-native/Rust memory outside the V8 heap). The default of `2` (→ 1 worker, 2
-processes total) is sized for the 16 GB / 4 vCPU GitHub-hosted runners the
-publish pipeline uses. At `8` (→ 7 workers) that runner ran out of memory and
-buildkit failed the step with `ResourceExhausted: ... cannot allocate memory`;
-`3` (→ 2 workers) still didn't fit once the per-process RSS was measured
-directly instead of inferred. `tests/unit/docker-build-memory-budget.test.ts`
-does the arithmetic against the measured figure and fails if either knob
-outgrows the runner. Both images (node and Bun) share these defaults; the Bun
-image's are set in `Dockerfile.bun` (Turbopack on Bun 1.4+, `#11719`).
+page-data worker is its own process and inherits `NODE_OPTIONS`, so the heap
+ceiling is per process, not per build: the default of `3` (→ 2 workers) is sized
+for the 16 GB / 4 vCPU GitHub-hosted runners the publish pipeline uses. At `8`
+(→ 7 workers) that runner ran out of memory and buildkit failed the step with
+`ResourceExhausted: ... cannot allocate memory`. `tests/unit/docker-build-memory-budget.test.ts`
+does the arithmetic and fails if either knob outgrows the runner.

 Turbopack compiles in native Rust memory that lives **outside** the V8 heap, so
 `OMNIROUTE_BUILD_MEMORY_MB` does not bound it. On a host with a memory ceiling the
--- a/docs/i18n/ar/llm.txt
+++ b/docs/i18n/ar/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/az/llm.txt
+++ b/docs/i18n/az/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/bg/llm.txt
+++ b/docs/i18n/bg/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/bn/llm.txt
+++ b/docs/i18n/bn/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/cs/llm.txt
+++ b/docs/i18n/cs/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/da/llm.txt
+++ b/docs/i18n/da/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/de/llm.txt
+++ b/docs/i18n/de/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/es/llm.txt
+++ b/docs/i18n/es/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/fa/llm.txt
+++ b/docs/i18n/fa/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/fi/llm.txt
+++ b/docs/i18n/fi/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/fr/llm.txt
+++ b/docs/i18n/fr/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/gu/llm.txt
+++ b/docs/i18n/gu/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/he/llm.txt
+++ b/docs/i18n/he/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/hi/llm.txt
+++ b/docs/i18n/hi/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/hu/llm.txt
+++ b/docs/i18n/hu/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/id/llm.txt
+++ b/docs/i18n/id/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/in/llm.txt
+++ b/docs/i18n/in/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/it/llm.txt
+++ b/docs/i18n/it/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/ja/llm.txt
+++ b/docs/i18n/ja/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/ko/llm.txt
+++ b/docs/i18n/ko/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/mr/llm.txt
+++ b/docs/i18n/mr/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/ms/llm.txt
+++ b/docs/i18n/ms/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/nl/llm.txt
+++ b/docs/i18n/nl/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/no/llm.txt
+++ b/docs/i18n/no/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/phi/llm.txt
+++ b/docs/i18n/phi/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/pl/llm.txt
+++ b/docs/i18n/pl/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/pt-BR/llm.txt
+++ b/docs/i18n/pt-BR/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/pt/llm.txt
+++ b/docs/i18n/pt/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/ro/llm.txt
+++ b/docs/i18n/ro/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/ru/llm.txt
+++ b/docs/i18n/ru/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/sk/llm.txt
+++ b/docs/i18n/sk/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/sv/llm.txt
+++ b/docs/i18n/sv/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/sw/llm.txt
+++ b/docs/i18n/sw/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/ta/llm.txt
+++ b/docs/i18n/ta/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/te/llm.txt
+++ b/docs/i18n/te/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/th/llm.txt
+++ b/docs/i18n/th/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/tr/llm.txt
+++ b/docs/i18n/tr/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/uk-UA/llm.txt
+++ b/docs/i18n/uk-UA/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/ur/llm.txt
+++ b/docs/i18n/ur/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/vi/llm.txt
+++ b/docs/i18n/vi/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/zh-CN/llm.txt
+++ b/docs/i18n/zh-CN/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/i18n/zh-TW/llm.txt
+++ b/docs/i18n/zh-TW/llm.txt
@@ -18,7 +18,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -438,7 +438,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/docs/ops/RELEASE_CHECKLIST.md
+++ b/docs/ops/RELEASE_CHECKLIST.md
@@ -1,12 +1,12 @@
 ---
 title: "Release Checklist"
-version: 3.8.51
-lastUpdated: 2026-08-28
+version: 3.8.40
+lastUpdated: 2026-06-28
 ---

 # Release Checklist

-> **Last updated:** 2026-08-28 — v3.8.51
+> **Last updated:** 2026-06-28 — v3.8.40
 > Streamlined release flow that leverages Claude Code skills for automation.
 >
 > **Keep the queue/branch green between releases:** see [RELEASE_GREEN.md](./RELEASE_GREEN.md)
@@ -37,21 +37,7 @@ npm run test:e2e           # optional but recommended
 /capture-release-evidences-cc
 ```

-## npm Trusted Publishing (default since v3.8.51) — staged on request, direct as fallback
-
-`npm-publish.yml` publishes through **npm Trusted Publishing (OIDC)** by default: the
-`stage-npm` job (github-hosted) exchanges GitHub's id-token for a short-lived npm
-credential for that run — no long-lived npm token in the repository secrets, no 2FA prompt, provenance attached.
-That is the bypass npm sanctions now that tokens which skip 2FA are being retired;
-it restores the fully automatic flow the project had up to v3.8.48 while keeping the
-WS1.3 guarantee (a leaked token cannot publish alone — there is no token).
-
-**One-time setup (owner):** npmjs.com → package `omniroute` → Settings → *Trusted
-Publisher* → GitHub: owner `diegosouzapw`, repo `OmniRoute`, workflow `npm-publish.yml`
-(environment: none). Until that exists, the automatic step fails with `ENEEDAUTH`:
-re-dispatch with `publish_mode=staged` (below) or `direct`.
-
-### Staged publishing (on request — `publish_mode=staged`)
+## npm Staged Publishing (default since v3.8.49 — WS1.3/D2)

 The npm-publish workflow no longer publishes directly: it boots the packed tarball
 (`check:pack-boot`) and then runs `npm stage publish` — the exact bytes are parked on
--- a/docs/ops/RUNNER_BOX.md
+++ b/docs/ops/RUNNER_BOX.md
@@ -4,66 +4,32 @@ title: Self-Hosted Runner Box Operations

 # Self-Hosted Runner Box Operations (.113 pool)

-The self-hosted pool (`self-hosted, omni-release` on all eight runners; `omni-build` on two) runs on the **.113** box.
-Measured 2026-08-28 (v3.8.50 postmortem, Parte III):
+The self-hosted pool (`self-hosted, omni-release` labels) runs on the 16 GB box at
+`192.168.0.113`. Two failure modes recurred on release days and were, until v3.8.49,
+manual discipline; the **janitor script codifies them** (WS3.3 of the quality plan):

-| resource  | value                                                            | what it means for scheduling                                                                                                                  |
-| --------- | ---------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------- |
-| RAM / CPU | **31 GB / 32 cores** (was 16 GB when this doc was first written) | one `next-build` peaks at **~14 GB** → 2 concurrent heavy builds saturate the box, 3 take it down (2026-08-28 06:42Z: load 56, two jobs lost) |
-| swap      | 15 GB                                                            | it swapped its way through the v3.8.50 publish; pressure shows in `/proc/pressure/memory`                                                     |
-| `/tmp`    | **12 GB tmpfs = RAM**                                            | anything parked there is memory; leftovers are swept after 3 h                                                                                |
-| disk      | 188 GB                                                           | `_work` checkouts of 8 runners reach ~70 GB with no cap                                                                                       |
-| runners   | **10 listeners**: 8 OmniRoute + OmniHeuris + OmniMind            | all share the memory above                                                                                                                    |
+1. **Orphaned temp/work dirs** filling the disk → disk-full SQLite errors mid-job.
+2. **>4 concurrent runners** → OOM-killed jobs (8-wide killed jobs twice on the
+   v3.8.47 release day; 4-wide is the proven ceiling).

 ## Install the janitor (one-time, on the box)

 ```bash
-scp scripts/ops/runner-janitor.sh root@192.168.0.113:/opt/omniroute-ops/runner-janitor.sh
-ssh root@192.168.0.113 'chmod +x /opt/omniroute-ops/runner-janitor.sh; apt-get install -y lsof'
-# cron (root): every 30 min, log to /var/log/runner-janitor.log
-*/30 * * * * MAX_ACTIVE_RUNNERS=8 /opt/omniroute-ops/runner-janitor.sh >> /var/log/runner-janitor.log 2>&1
+sudo mkdir -p /opt/omniroute-ops
+sudo cp scripts/ops/runner-janitor.sh /opt/omniroute-ops/
+sudo chmod +x /opt/omniroute-ops/runner-janitor.sh
+( sudo crontab -l 2>/dev/null; echo '*/30 * * * * /opt/omniroute-ops/runner-janitor.sh >> /var/log/runner-janitor.log 2>&1' ) | sudo crontab -
 ```

-`lsof` is required: the janitor proves a path is idle with one snapshot of open
-files before removing it, and without the tool it removes nothing and says so
-(exit 1). Try any change with `--dry-run` first — it prints exactly what it would
-do and touches nothing.
-
-What it does every run: sweeps our own leftovers (`runner-*`, `omniroute-*`,
-`next-build*`, `e2e-build.tar.gz`) after **3 h on tmpfs** and 24 h on disk
-`_work/_temp`; kills a `next-build` older than 75 min (no job runs that long — on
-2026-08-27 one ran 70 min after GitHub had declared its job lost); prunes 48 h-old
-checkouts of runners whose unit is **stopped**; alerts on disk ≥ 85 %, memory PSI
-`full/avg60` ≥ 10 %, and more listeners than `MAX_ACTIVE_RUNNERS` (with an
-omniroute/other breakdown). Exit 1 = attention needed; read the log.
-
-## Runner units: KillMode
-
-The runner's default `KillMode=process` leaves `Runner.Worker → npm → next-build`
-alive when a unit is stopped or restarted — an orphan build keeps eating RAM and
-CPU with no job attached. Every OmniRoute unit carries a drop-in
-(`/etc/systemd/system/actions.runner.diegosouzapw-OmniRoute.<name>.service.d/10-killmode.conf`)
-with `KillMode=mixed`: SIGTERM to the listener first, SIGKILL to the whole cgroup at
-`TimeoutStop`. It takes effect on the unit's next restart — restart **one runner at
-a time, only when idle**, with the idle check and the restart in the same command.
+What it does every 30min: sweeps runner temp leftovers older than 24h, alerts at
+≥85% root-disk usage, and alerts when more than the runner ceiling (default 4, tunable
+via the script's own environment) of `Runner.Listener` processes are up. Alerts land in `/var/log/runner-janitor.log`
+with a non-zero exit (grep for `⚠`).

 ## Operating rules

- **Heavy-build ceiling: 2 at a time — enforced by label.** Every job that runs a
-  `next build` (`ci.yml` `build`, `npm-publish.yml` `publish`, both `nightly-release-green`
-  validations) targets `[self-hosted, omni-build]`, and only **two** runners carry that
-  label (`omniroute-113-5`, `omniroute-113-6`, added through the runners API — no
-  re-registration). The other six keep `omni-release` and take nothing heavy; GitHub
-  queues a third build instead of the kernel killing one. Pair with the `heavy-build-*`
-  concurrency lanes in `ci.yml`. To add capacity, label another runner — never raise
-  the count past what 31 GB holds (one next-build ≈ 14–16 GB).
- **Never clean `/tmp` or `_work` by hand while any runner is busy.** A
-  check-then-delete with a gap between the two is how a live Build job lost its
-  `_work` on 2026-08-27. The janitor does the check and the removal in one step;
-  let it.
- Stopping a runner mid-job cancels the job (observed live): `systemctl stop` only
-  when its listener has no `Runner.Worker` child — and do it in one command.
- Workflows must not park artefacts in `/tmp` (it is RAM). Download to
-  `$RUNNER_TEMP` (on disk, per runner) — the 1.3 GB `next-build` artefact took 27–32
-  minutes to land on the tmpfs and 2 minutes to upload from disk.
+- **Ceiling: 4 runners** on the 16 GB box. Runners 5–8 stay STOPPED except for
+  explicit off-peak experiments — never during a release window.
+- Stopping a runner mid-job cancels the job (observed live): `systemctl stop`
+  only when its runner is idle (`Runner.Listener` without a `Runner.Worker` child).
 - The `.15` VPS is homologation-only — never runs CI runners.
--- a/docs/reference/ENVIRONMENT.md
+++ b/docs/reference/ENVIRONMENT.md
@@ -103,7 +103,6 @@ OmniRoute uses **SQLite** (via `better-sqlite3`) for all persistence. These vari
 | `OMNIROUTE_MIGRATIONS_DIR`             | _(auto-detect)_      | `src/lib/db/migrationRunner.ts`                       | Override the directory that the migration runner scans. Useful when shipping bundled migrations in custom builds.                                                                                                                               |
 | `OMNIROUTE_EXTRA_MIGRATIONS_DIRS`      | _(unset)_            | `src/lib/db/migrationRunner/extraDirs.ts`             | Additional migration directories as `namespace=dir` entries separated by the platform path delimiter (e.g. `ee=/opt/app/enterprise/db/migrations`). Files found there are recorded as `<namespace>-<number>`, so a distribution shipping its own migrations never collides with the upstream numeric slots. A malformed entry, an invalid namespace or a missing directory throws at startup instead of silently skipping the schema. |
 | `OMNIROUTE_MAX_PENDING_MIGRATIONS`     | `50`                 | `src/lib/db/migrationRunner.ts`                       | Mass-pending-migrations safety threshold (#3416). Startup aborts if more than this many migrations are pending on an existing DB (guards against a wiped tracking table). Raise it to restore an older backup; set to `0` to disable the check. |
-| `OMNIROUTE_INSTALL_UPGRADE_WORKDIR`     | _(`<repo>/.install-upgrade`)_ | `scripts/check/check-install-upgrade.mjs`             | Working directory for the `check:install-upgrade` release gate. It needs roughly 12 GB (two ~3 GB install trees plus the tarball), so it must not run on a small tmpfs — on the self-hosted runner `/tmp` is a 12 GB RAM-backed tmpfs and the gate exhausted it, truncating the package. |
 | `OMNIROUTE_SPEND_FLUSH_INTERVAL_MS`    | _(default in code)_  | `src/lib/spend/batchWriter.ts`                        | Flush interval (ms) for the batched spend/cost writer. Lower values reduce write coalescing; higher values reduce DB contention.                                                                                                                |
 | `OMNIROUTE_SPEND_MAX_BUFFER_SIZE`      | _(default in code)_  | `src/lib/spend/batchWriter.ts`                        | Max buffered spend entries before a forced flush. Raise on high-QPS deployments; lower when bounded memory matters more.                                                                                                                        |
 | `OMNIROUTE_PROXY_FETCH_DEBUG`          | _(unset)_            | `open-sse/utils/proxyFetch.ts`                        | Set to `"true"` to emit `[ProxyFetch]` debug logs on the Vercel relay path. Off by default to avoid leaking routing hints.                                                                                                                      |
--- a/electron/package-lock.json
+++ b/electron/package-lock.json
@@ -297,45 +297,6 @@
        "url": "https://github.com/sponsors/isaacs"
      }
    },
-    "node_modules/@electron/windows-sign": {
-      "version": "1.2.2",
-      "resolved": "https://registry.npmjs.org/@electron/windows-sign/-/windows-sign-1.2.2.tgz",
-      "integrity": "sha512-dfZeox66AvdPtb2lD8OsIIQh12Tp0GNCRUDfBHIKGpbmopZto2/A8nSpYYLoedPIHpqkeblZ/k8OV0Gy7PYuyQ==",
-      "dev": true,
-      "license": "BSD-2-Clause",
-      "optional": true,
-      "peer": true,
-      "dependencies": {
-        "cross-dirname": "^0.1.0",
-        "debug": "^4.3.4",
-        "fs-extra": "^11.1.1",
-        "minimist": "^1.2.8",
-        "postject": "^1.0.0-alpha.6"
-      },
-      "bin": {
-        "electron-windows-sign": "bin/electron-windows-sign.js"
-      },
-      "engines": {
-        "node": ">=14.14"
-      }
-    },
-    "node_modules/@electron/windows-sign/node_modules/fs-extra": {
-      "version": "11.4.0",
-      "resolved": "https://registry.npmjs.org/fs-extra/-/fs-extra-11.4.0.tgz",
-      "integrity": "sha512-EQsFzMUJkCKGr1ePqlYADkIUmHW1s3ZXr5Yqy6wbGrfUCphpl2maM/kyOIRA2HpP3AaFQTZXD4ldjek+nccddA==",
-      "dev": true,
-      "license": "MIT",
-      "optional": true,
-      "peer": true,
-      "dependencies": {
-        "graceful-fs": "^4.2.0",
-        "jsonfile": "^6.0.1",
-        "universalify": "^2.0.0"
-      },
-      "engines": {
-        "node": ">=14.14"
-      }
-    },
    "node_modules/@isaacs/fs-minipass": {
      "version": "4.0.1",
      "resolved": "https://registry.npmjs.org/@isaacs/fs-minipass/-/fs-minipass-4.0.1.tgz",
@@ -1130,15 +1091,6 @@
      "dev": true,
      "license": "MIT"
    },
-    "node_modules/cross-dirname": {
-      "version": "0.1.0",
-      "resolved": "https://registry.npmjs.org/cross-dirname/-/cross-dirname-0.1.0.tgz",
-      "integrity": "sha512-+R08/oI0nl3vfPcqftZRpytksBXDzOUveBq/NBVx0sUp1axwzPQrKinNx5yd5sxPu8j1wIy8AfnVQ+5eFdha6Q==",
-      "dev": true,
-      "license": "MIT",
-      "optional": true,
-      "peer": true
-    },
    "node_modules/cross-spawn": {
      "version": "7.0.6",
      "resolved": "https://registry.npmjs.org/cross-spawn/-/cross-spawn-7.0.6.tgz",
@@ -1459,19 +1411,6 @@
        "node": ">=14.0.0"
      }
    },
-    "node_modules/electron-builder-squirrel-windows": {
-      "version": "26.15.3",
-      "resolved": "https://registry.npmjs.org/electron-builder-squirrel-windows/-/electron-builder-squirrel-windows-26.15.3.tgz",
-      "integrity": "sha512-Jc19XPV9y9+2bAdZPkXuVNGNIEFBq9poHC61l8Kv6FdK7DRG3+Ic0rerC0DXOaeHNz8yW0fg/JnF8GQROOF5MA==",
-      "dev": true,
-      "license": "MIT",
-      "peer": true,
-      "dependencies": {
-        "app-builder-lib": "26.15.3",
-        "builder-util": "26.15.3",
-        "electron-winstaller": "5.4.0"
-      }
-    },
    "node_modules/electron-publish": {
      "version": "26.15.3",
      "resolved": "https://registry.npmjs.org/electron-publish/-/electron-publish-26.15.3.tgz",
@@ -1506,66 +1445,6 @@
        "tiny-typed-emitter": "^2.1.0"
      }
    },
-    "node_modules/electron-winstaller": {
-      "version": "5.4.0",
-      "resolved": "https://registry.npmjs.org/electron-winstaller/-/electron-winstaller-5.4.0.tgz",
-      "integrity": "sha512-bO3y10YikuUwUuDUQRM4KfwNkKhnpVO7IPdbsrejwN9/AABJzzTQ4GeHwyzNSrVO+tEH3/Np255a3sVZpZDjvg==",
-      "dev": true,
-      "hasInstallScript": true,
-      "license": "MIT",
-      "peer": true,
-      "dependencies": {
-        "@electron/asar": "^3.2.1",
-        "debug": "^4.1.1",
-        "fs-extra": "^7.0.1",
-        "lodash": "^4.17.21",
-        "temp": "^0.9.0"
-      },
-      "engines": {
-        "node": ">=8.0.0"
-      },
-      "optionalDependencies": {
-        "@electron/windows-sign": "^1.1.2"
-      }
-    },
-    "node_modules/electron-winstaller/node_modules/fs-extra": {
-      "version": "7.0.1",
-      "resolved": "https://registry.npmjs.org/fs-extra/-/fs-extra-7.0.1.tgz",
-      "integrity": "sha512-YJDaCJZEnBmcbw13fvdAM9AwNOJwOzrE4pqMqBq5nFiEqXUqHwlK4B+3pUw6JNvfSPtX05xFHtYy/1ni01eGCw==",
-      "dev": true,
-      "license": "MIT",
-      "peer": true,
-      "dependencies": {
-        "graceful-fs": "^4.1.2",
-        "jsonfile": "^4.0.0",
-        "universalify": "^0.1.0"
-      },
-      "engines": {
-        "node": ">=6 <7 || >=8"
-      }
-    },
-    "node_modules/electron-winstaller/node_modules/jsonfile": {
-      "version": "4.0.0",
-      "resolved": "https://registry.npmjs.org/jsonfile/-/jsonfile-4.0.0.tgz",
-      "integrity": "sha512-m6F1R3z8jjlf2imQHS2Qez5sjKWQzbuuhuJ/FKYFRZvPE3PuHcSMVZzfsLhGVOkfd20obL5SWEBew5ShlquNxg==",
-      "dev": true,
-      "license": "MIT",
-      "peer": true,
-      "optionalDependencies": {
-        "graceful-fs": "^4.1.6"
-      }
-    },
-    "node_modules/electron-winstaller/node_modules/universalify": {
-      "version": "0.1.2",
-      "resolved": "https://registry.npmjs.org/universalify/-/universalify-0.1.2.tgz",
-      "integrity": "sha512-rBJeI5CXAlmy1pV+617WB9J63U6XcazHHF2f2dbJix4XzpUF0RS3Zbj0FGIOCAva5P/d/GBOYaACQ1w+0azUkg==",
-      "dev": true,
-      "license": "MIT",
-      "peer": true,
-      "engines": {
-        "node": ">= 4.0.0"
-      }
-    },
    "node_modules/emoji-regex": {
      "version": "8.0.0",
      "resolved": "https://registry.npmjs.org/emoji-regex/-/emoji-regex-8.0.0.tgz",
@@ -2480,20 +2359,6 @@
        "node": ">= 18"
      }
    },
-    "node_modules/mkdirp": {
-      "version": "0.5.6",
-      "resolved": "https://registry.npmjs.org/mkdirp/-/mkdirp-0.5.6.tgz",
-      "integrity": "sha512-FP+p8RB8OWpF3YZBCrP5gtADmtXApB5AMLn+vdyA+PyxCjrCs00mjyUozssO33cwDeT3wNGdLxJ5M//YqtHAJw==",
-      "dev": true,
-      "license": "MIT",
-      "peer": true,
-      "dependencies": {
-        "minimist": "^1.2.6"
-      },
-      "bin": {
-        "mkdirp": "bin/cmd.js"
-      }
-    },
    "node_modules/ms": {
      "version": "2.1.3",
      "resolved": "https://registry.npmjs.org/ms/-/ms-2.1.3.tgz",
@@ -2757,36 +2622,6 @@
        "node": ">=18"
      }
    },
-    "node_modules/postject": {
-      "version": "1.0.0-alpha.6",
-      "resolved": "https://registry.npmjs.org/postject/-/postject-1.0.0-alpha.6.tgz",
-      "integrity": "sha512-b9Eb8h2eVqNE8edvKdwqkrY6O7kAwmI8kcnBv1NScolYJbo59XUF0noFq+lxbC1yN20bmC0WBEbDC5H/7ASb0A==",
-      "dev": true,
-      "license": "MIT",
-      "optional": true,
-      "peer": true,
-      "dependencies": {
-        "commander": "^9.4.0"
-      },
-      "bin": {
-        "postject": "dist/cli.js"
-      },
-      "engines": {
-        "node": ">=14.0.0"
-      }
-    },
-    "node_modules/postject/node_modules/commander": {
-      "version": "9.5.0",
-      "resolved": "https://registry.npmjs.org/commander/-/commander-9.5.0.tgz",
-      "integrity": "sha512-KRs7WVDKg86PWiuAqhDrAQnTXZKraVcCc6vFdL14qrZ/DcWwuRo7VoiYXalXO7S5GKpqYiVEwCbgFDfxNHKJBQ==",
-      "dev": true,
-      "license": "MIT",
-      "optional": true,
-      "peer": true,
-      "engines": {
-        "node": "^12.20.0 || >=14"
-      }
-    },
    "node_modules/proc-log": {
      "version": "6.1.0",
      "resolved": "https://registry.npmjs.org/proc-log/-/proc-log-6.1.0.tgz",
@@ -2981,21 +2816,6 @@
        "node": ">= 4"
      }
    },
-    "node_modules/rimraf": {
-      "version": "2.6.3",
-      "resolved": "https://registry.npmjs.org/rimraf/-/rimraf-2.6.3.tgz",
-      "integrity": "sha512-mwqeW5XsA2qAejG46gYdENaxXjx9onRNCfn7L0duuP4hCuTIi/QO7PDK07KJfp1d+izWPrzEJDcSqBa0OZQriA==",
-      "deprecated": "Rimraf versions prior to v4 are no longer supported",
-      "dev": true,
-      "license": "ISC",
-      "peer": true,
-      "dependencies": {
-        "glob": "^7.1.3"
-      },
-      "bin": {
-        "rimraf": "bin.js"
-      }
-    },
    "node_modules/roarr": {
      "version": "2.15.4",
      "resolved": "https://registry.npmjs.org/roarr/-/roarr-2.15.4.tgz",
@@ -3251,21 +3071,6 @@
        "node": ">=18"
      }
    },
-    "node_modules/temp": {
-      "version": "0.9.4",
-      "resolved": "https://registry.npmjs.org/temp/-/temp-0.9.4.tgz",
-      "integrity": "sha512-yYrrsWnrXMcdsnu/7YMYAofM1ktpL5By7vZhf15CrXijWWrEYZks5AXBudalfSWJLlnen/QUJUB5aoB0kqZUGA==",
-      "dev": true,
-      "license": "MIT",
-      "peer": true,
-      "dependencies": {
-        "mkdirp": "^0.5.1",
-        "rimraf": "~2.6.2"
-      },
-      "engines": {
-        "node": ">=6.0.0"
-      }
-    },
    "node_modules/temp-file": {
      "version": "3.4.0",
      "resolved": "https://registry.npmjs.org/temp-file/-/temp-file-3.4.0.tgz",
--- a/llm.txt
+++ b/llm.txt
@@ -14,7 +14,7 @@ OmniRoute solves the problem of managing multiple AI provider subscriptions, quo

 - **Runtime:** Node.js `>=22.0.0 <23 || >=24.0.0 <27`, ES Modules (`"type": "module"`)
 - **Framework:** Next.js 16 (App Router) with TypeScript 6
- **Database:** SQLite via better-sqlite3 (local, zero-config, 160 migrations)
+- **Database:** SQLite via better-sqlite3 (local, zero-config, 159 migrations)
 - **State management:** Zustand (client), SQLite (server persistence)
 - **UI:** React 19, Tailwind CSS 4, Recharts for analytics, @lobehub/icons for 130+ provider SVG icons
 - **Auth:** OAuth 2.0 (PKCE) for providers, bcrypt for local user auth
@@ -434,7 +434,7 @@ diagnostics) plus **memory**, **skill**, **agentSkill**, **githubSkill**, **pool

 4. **Environment variables:** All configuration is in `.env` (from `.env.example`). Key vars: `PORT`, `NEXT_PUBLIC_BASE_URL`, `API_KEY`, `ADMIN_PASSWORD`.

-5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 160 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.
+5. **Database layer:** Operations go through `src/lib/db/` modules (120 domain-specific files, 159 migrations). `localDb.ts` is re-exports only — add new functions to the proper `db/*.ts` module.

 6. **Tests** use Node.js built-in test runner + Vitest. Run `npm test`. Vitest for MCP/autoCombo (`npm run test:vitest`). Playwright for E2E (`npm run test:e2e`). Coverage gate: ratchet vs `quality-baseline.json`, absolute floor 60% statements/lines/functions/branches.

--- a/package.json
+++ b/package.json
@@ -52,8 +52,7 @@
    "!**/*.test.js",
    "!**/*.test.mjs",
    "!**/*.spec.ts",
-    "!**/*.spec.tsx",
-    "!**/*.nft.json"
+    "!**/*.spec.tsx"
  ],
  "workspaces": [
    "open-sse",
--- a/scripts/ad-hoc/codemod-rm-maxretries.mjs
+++ b/scripts/ad-hoc/codemod-rm-maxretries.mjs
@@ -1,101 +0,0 @@
-#!/usr/bin/env node
-/**
- * One-shot codemod (#11966): give every recursive temp-dir removal in tests the retry
- * options Node already supports, so a WAL/backup/worker still writing into the directory
- * turns into a retried delete instead of a red shard:
- *
- *   rmSync(dir, { recursive: true, force: true })
- *   → rmSync(dir, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 })
- *
- * Applies to `rmSync(`, `fs.rmSync(`, `rm(` / `fs.rm(` / `fs.promises.rm(` (async) and
- * `rmdirSync(` calls whose option object literal contains `recursive: true` and no
- * `maxRetries`. Only the option object is touched — call sites, assertions and imports are
- * left as they are. Usage: node scripts/ad-hoc/codemod-rm-maxretries.mjs [dir=tests]
- */
-import fs from "node:fs";
-import path from "node:path";
-
-const root = process.argv[2] || "tests";
-const CALL = /\b(?:fs\.promises\.|fsp\.|fs\.|promises\.)?(?:rmSync|rmdirSync|rm)\(/g;
-let files = 0;
-let sites = 0;
-
-function walk(dir, out = []) {
-  for (const e of fs.readdirSync(dir, { withFileTypes: true })) {
-    const p = path.join(dir, e.name);
-    if (e.isDirectory()) {
-      if (e.name === "node_modules" || e.name === "fixtures") continue;
-      walk(p, out);
-    } else if (/\.(ts|tsx|mts|cts|js|mjs|cjs)$/.test(e.name)) out.push(p);
-  }
-  return out;
-}
-
-// Find the closing brace of the option object literal that starts at `open`.
-function objectEnd(src, open) {
-  let depth = 0;
-  for (let i = open; i < src.length; i++) {
-    const c = src[i];
-    if (c === "{") depth++;
-    else if (c === "}") {
-      depth--;
-      if (depth === 0) return i;
-    } else if (c === '"' || c === "'" || c === "`") {
-      const q = c;
-      i++;
-      while (i < src.length && src[i] !== q) {
-        if (src[i] === "\\") i++;
-        i++;
-      }
-    }
-  }
-  return -1;
-}
-
-for (const file of walk(root)) {
-  const src = fs.readFileSync(file, "utf8");
-  let out = "";
-  let last = 0;
-  let touched = 0;
-  for (const m of src.matchAll(CALL)) {
-    const callStart = m.index + m[0].length;
-    // Locate the option object: the first `{` before the call's closing paren at depth 0.
-    let depth = 0;
-    let objOpen = -1;
-    for (let i = callStart; i < src.length; i++) {
-      const c = src[i];
-      if (c === "(" || c === "[") depth++;
-      else if (c === ")" || c === "]") {
-        if (depth === 0) break;
-        depth--;
-      } else if (c === "{" && depth === 0) {
-        objOpen = i;
-        break;
-      }
-    }
-    if (objOpen === -1) continue;
-    const objClose = objectEnd(src, objOpen);
-    if (objClose === -1) continue;
-    const obj = src.slice(objOpen, objClose + 1);
-    if (!/\brecursive:\s*true\b/.test(obj) || /\bmaxRetries\b/.test(obj)) continue;
-    // Insert before the closing brace, respecting an existing trailing comma / newline.
-    const inner = obj.slice(1, -1);
-    const trimmed = inner.replace(/\s+$/, "");
-    const trailing = inner.slice(trimmed.length);
-    const sep = trimmed.endsWith(",") ? " " : ", ";
-    const multiline = /\n/.test(trailing);
-    const insert = multiline
-      ? `${trimmed}${trimmed.endsWith(",") ? "" : ","}\n${trailing.replace(/\n$/, "")}  maxRetries: 5,\n  retryDelay: 100,${trailing}`
-      : `${trimmed}${sep}maxRetries: 5, retryDelay: 100${trailing}`;
-    out += src.slice(last, objOpen + 1) + insert;
-    last = objClose;
-    touched++;
-  }
-  if (touched) {
-    out += src.slice(last);
-    fs.writeFileSync(file, out);
-    files++;
-    sites += touched;
-  }
-}
-console.log(`[codemod-rm-maxretries] ${sites} call site(s) in ${files} file(s) under ${root}`);
--- a/scripts/build/build-next-isolated.mjs
+++ b/scripts/build/build-next-isolated.mjs
@@ -131,15 +131,9 @@ function runNextBuild() {
 }

 export function resolveNextBuildBundlerFlag(baseEnv = process.env) {
-  // Turbopack is the default; OMNIROUTE_USE_TURBOPACK=0 is the documented escape hatch
-  // to webpack (Windows, native-binding trouble, RAM-constrained machines — #6409, and
-  // docs/reference/ENVIRONMENT.md). The choice is env-only ON PURPOSE: the variable is
-  // the operator's control and CI sets it explicitly, so sniffing the runtime here would
-  // silently override an operator who asked for Turbopack. Bun 1.4+ supports Turbopack's
-  // V8 worker bindings (#11471), so the historical `process.versions.bun` → `--webpack`
-  // hardcode is gone; the `OMNIROUTE_USE_TURBOPACK=0` fallback remains for Bun < 1.4
-  // images built with the webpack path.
-  if (baseEnv.OMNIROUTE_USE_TURBOPACK === "0") {
+  // Turbopack is the default on Node.js; on Bun or when explicitly disabled (=0),
+  // use Webpack (--webpack) to avoid Turbopack V8 internal worker API mismatches.
+  if (process.versions.bun || baseEnv.OMNIROUTE_USE_TURBOPACK === "0") {
    return "--webpack";
  }
  return "--turbopack";
--- a/scripts/check/check-install-upgrade.mjs
+++ b/scripts/check/check-install-upgrade.mjs
@@ -30,12 +30,10 @@
 * Requires `npm run build:cli` first — this is a --with-build gate, like check:pack-boot.
 */

-import { execFileSync, spawn, spawnSync } from "node:child_process";
-import crypto from "node:crypto";
+import { execFileSync, spawn } from "node:child_process";
 import fs from "node:fs";
 import os from "node:os";
 import path from "node:path";
-import { pathToFileURL } from "node:url";
 import { DatabaseSync } from "node:sqlite";

 const BOOT_DEADLINE_MS = 180_000;
@@ -45,59 +43,6 @@ const ALLOWLIST_PATH = "config/quality/install-upgrade-allowlist.json";
 const log = (msg) => console.log(`[install-upgrade] ${msg}`);
 const warn = (msg) => console.log(`[install-upgrade] ⚠️  ${msg}`);

-/** Root of the installed package inside an `npm install -g --prefix` tree. */
-function packageRootFor(prefix) {
-  return path.join(prefix, "lib", "node_modules", "omniroute");
-}
-
-/**
- * Credential for the health probe.
- *
- * GHSA-mvf8-qc78-5mxm hardened /api/monitoring/health: an ANONYMOUS caller now gets only
- * `{ status }` — the version, node version, pid and provider config are reserved for a
- * management principal (src/app/api/monitoring/health/route.ts → publicHealthView). An
- * unauthenticated probe therefore reads `body.version === undefined`, and this gate's
- * version assertion could never pass again; the v3.8.50 publish run failed with
- * "clean: health reports version undefined, expected 3.8.50" for exactly that reason.
- *
- * The gate spawns the server itself, so it can mint the credential instead of guessing one:
- * `OMNIROUTE_INTERNAL_SERVICE_TOKEN` + the `x-omniroute-internal-service-token` header is
- * accepted by requireManagementAuth() via isTrustedLoopbackInternalServiceRequest(), and the
- * probe is loopback by construction. Unlike the machine token `check:pack-boot` derives, this
- * does not depend on a readable machine-id, and an older PREVIOUS version that never gated
- * health simply ignores the header. The assertion keeps its full strength — it just presents
- * a credential.
- */
-const INTERNAL_SERVICE_TOKEN = crypto.randomBytes(32).toString("hex");
-const INTERNAL_SERVICE_HEADER = "x-omniroute-internal-service-token";
-
-/**
- * Secondary credential: the same loopback machine token `check:pack-boot` derives from the
- * packaged CLI. Sent alongside the internal-service token so a build that only honours one
- * of the two still answers with the full payload.
- */
-function derivePackagedCliToken(prefix) {
-  const cliModuleUrl = pathToFileURL(
-    path.join(packageRootFor(prefix), "bin", "cli", "utils", "cliToken.mjs")
-  ).href;
-  try {
-    return execFileSync(
-      process.execPath,
-      [
-        "--input-type=module",
-        "--eval",
-        "import(process.argv[1]).then(async m => process.stdout.write(await m.getCliToken()))",
-        cliModuleUrl,
-      ],
-      { encoding: "utf8", env: { ...process.env } }
-    ).trim();
-  } catch {
-    // A truncated/broken install cannot derive a token. Returning null keeps the boot
-    // probe running (it will fail loudly on its own) instead of crashing the gate here.
-    return null;
-  }
-}
-
 function pickTarball(packJson) {
  const filename = JSON.parse(packJson)?.[0]?.filename;
  if (!filename) throw new Error("npm pack --json returned no filename");
@@ -176,11 +121,6 @@ async function bootAndProbe({ prefix, dataDir, port, expectVersion, label }) {
  if (!fs.existsSync(binPath)) {
    return { ok: false, failures: [`${label}: bin not found at ${binPath}`], tail: [] };
  }
-  const cliToken = derivePackagedCliToken(prefix);
-  const probeHeaders = {
-    [INTERNAL_SERVICE_HEADER]: INTERNAL_SERVICE_TOKEN,
-    ...(cliToken ? { "x-omniroute-cli-token": cliToken } : {}),
-  };
  const child = spawn(binPath, ["serve", "--port", String(port)], {
    env: {
      ...process.env,
@@ -190,7 +130,6 @@ async function bootAndProbe({ prefix, dataDir, port, expectVersion, label }) {
      API_KEY_SECRET: "install-upgrade-gate-api-key-secret-long",
      DISABLE_SQLITE_AUTO_BACKUP: "true",
      OMNIROUTE_SKIP_SYSTEM_TRUST: "1",
-      OMNIROUTE_INTERNAL_SERVICE_TOKEN: INTERNAL_SERVICE_TOKEN,
    },
    stdio: ["ignore", "pipe", "pipe"],
    detached: true,
@@ -212,33 +151,20 @@ async function bootAndProbe({ prefix, dataDir, port, expectVersion, label }) {
  let result = { ok: false, failures: [`${label}: never became healthy`], tail };
  while (Date.now() < deadline) {
    if (childExit !== null) {
-      result = {
-        ok: false,
-        failures: [`${label}: exited with code ${childExit} before serving`],
-        tail,
-      };
+      result = { ok: false, failures: [`${label}: exited with code ${childExit} before serving`], tail };
      break;
    }
    try {
-      const res = await fetch(`http://127.0.0.1:${port}/api/monitoring/health`, {
-        headers: probeHeaders,
-      });
+      const res = await fetch(`http://127.0.0.1:${port}/api/monitoring/health`);
      const body = await res.json().catch(() => null);
      if (res.status === 200 && body && typeof body === "object") {
        const failures = [];
        // `status` may legitimately report degraded (no providers configured) — the gate
        // targets boot crashes and version mismatches, not health of a bare install.
-        const reportedVersion = body.version ?? body.system?.version;
-        if (expectVersion && reportedVersion !== expectVersion) {
-          failures.push(
-            `${label}: health reports version ${reportedVersion}, expected ${expectVersion}` +
-              (reportedVersion === undefined
-                ? " — the payload carries no version at all, which is the ANONYMOUS health " +
-                  "view: the probe's credentials were not accepted (see GHSA-mvf8-qc78-5mxm)"
-                : "")
-          );
+        if (expectVersion && body.version !== expectVersion) {
+          failures.push(`${label}: health reports version ${body.version}, expected ${expectVersion}`);
        }
-        result = { ok: failures.length === 0, version: reportedVersion, failures, tail };
+        result = { ok: failures.length === 0, version: body.version, failures, tail };
        break;
      }
    } catch {
@@ -257,72 +183,13 @@ async function bootAndProbe({ prefix, dataDir, port, expectVersion, label }) {
  return result;
 }

-/**
- * `npm install` reports ENOSPC as a *warning* per failed tar entry and still exits 0.
- *
- * That is not a theoretical concern: on the v3.8.50 publish run the Phase B upgrade install
- * emitted 5611 `npm warn tar TAR_ENTRY_ERROR ENOSPC: no space left on device` lines, exited
- * 0, and left a truncated package behind. `omniroute serve` then "exited with code 0 before
- * serving", no migration ever ran, and the gate concluded the release was missing 15 tables
- * — a full false alarm produced by a full disk. Each install tree is ~3 GB, and the run
- * builds two of them plus a ~275 MB tarball.
- *
- * So: surface the truncation at the install, where it is unambiguous.
- */
-function npmInstallInto(prefix, spec, label = spec) {
-  // spawnSync (not execFileSync): execFileSync forwards the child's stderr straight to the
-  // parent's, so the ENOSPC warnings scrolled past in CI without the script ever seeing
-  // them. spawnSync hands both streams back.
-  const run = spawnSync(
-    "npm",
-    ["install", "-g", "--prefix", prefix, "--no-audit", "--no-fund", spec],
-    { encoding: "utf8", maxBuffer: 512 * 1024 * 1024 }
-  );
-  const output = `${run.stdout ?? ""}${run.stderr ?? ""}`;
-  // Keep the install log visible, but a truncated package emits thousands of identical
-  // warnings — collapse them so the real message is not buried.
-  const stderrLines = String(run.stderr ?? "").split("\n");
-  const shown = stderrLines.length > 60 ? stderrLines.slice(0, 40) : stderrLines;
-  if (String(run.stderr ?? "").trim()) {
-    process.stderr.write(shown.join("\n") + "\n");
-    if (stderrLines.length > 60) {
-      process.stderr.write(`[install-upgrade] … ${stderrLines.length - 40} more npm line(s)\n`);
-    }
-  }
-  assertNoDiskExhaustion(output, label);
-  if (run.error) throw run.error;
-  if (run.status !== 0) {
-    throw new Error(`${label}: npm install exited with code ${run.status}`);
-  }
+function npmInstallInto(prefix, spec) {
+  execFileSync("npm", ["install", "-g", "--prefix", prefix, "--no-audit", "--no-fund", spec], {
+    encoding: "utf8",
+    maxBuffer: 128 * 1024 * 1024,
+  });
 }

-export function assertNoDiskExhaustion(output, label) {
-  if (!/ENOSPC|no space left on device/i.test(output)) return;
-  const count = (output.match(/ENOSPC/g) ?? []).length;
-  throw new Error(
-    `${label}: the install ran out of disk space (${count} ENOSPC error(s) from npm). ` +
-      `The package tree is truncated, so anything measured from it — boot, schema, ` +
-      `migrations — is meaningless. Free space in ${workDirForMessages} (each install tree is ` +
-      `~3 GB) and re-run. This is an environment failure, NOT a schema divergence.`
-  );
-}
-
-/** Best-effort free bytes on the filesystem backing `dir`, or null when unavailable. */
-function freeBytes(dir) {
-  try {
-    return fs.statfsSync(dir).bavail * fs.statfsSync(dir).bsize;
-  } catch {
-    return null;
-  }
-}
-
-const GB = 1024 ** 3;
-
-// Set once the work directory exists, so the ENOSPC message names the filesystem that
-// actually ran out — pointing at /tmp when the gate works elsewhere sends the reader to
-// free space on the wrong volume (which is what happened during the v3.8.50 publish).
-let workDirForMessages = os.tmpdir();
-
 function resolvePreviousVersion(current, explicit) {
  if (explicit) return explicit;
  const out = execFileSync("npm", ["view", "omniroute", "dist-tags.latest"], { encoding: "utf8" });
@@ -331,9 +198,7 @@ function resolvePreviousVersion(current, explicit) {
  if (latest === current) {
    // The version under test is already published (re-run of a shipped release): step back
    // to the highest published version strictly below it.
-    const all = JSON.parse(
-      execFileSync("npm", ["view", "omniroute", "versions", "--json"], { encoding: "utf8" })
-    );
+    const all = JSON.parse(execFileSync("npm", ["view", "omniroute", "versions", "--json"], { encoding: "utf8" }));
    const stable = all.filter((v) => !/-(rc|alpha|beta|pre|next)/.test(v) && v !== current);
    return stable[stable.length - 1];
  }
@@ -353,25 +218,11 @@ async function main() {
  }
  const version = JSON.parse(fs.readFileSync(path.join(ROOT, "package.json"), "utf8")).version;
  const allowlist = loadAllowlist(ROOT);
-  // NOT os.tmpdir(): on the self-hosted runner /tmp is a 12 GB tmpfs backed by RAM, while
-  // the root filesystem has ~66 GB free. This gate needs ~12 GB, so it exhausted the tmpfs
-  // and npm truncated the package — 58269 ENOSPC errors on the v3.8.50 publish, which the
-  // previous code could only report as a crash. Freeing disk did not help because the disk
-  // was never the constraint. Work on real disk beside the repo instead.
-  const workRoot =
-    process.env.OMNIROUTE_INSTALL_UPGRADE_WORKDIR || path.join(ROOT, ".install-upgrade");
-  fs.mkdirSync(workRoot, { recursive: true });
-  const tmp = fs.mkdtempSync(path.join(workRoot, "omniroute-install-upgrade-"));
-  workDirForMessages = tmp;
+  const tmp = fs.mkdtempSync(path.join(os.tmpdir(), "omniroute-install-upgrade-"));
  const failures = [];
  const warnings = [];

  try {
-    // Timed, because this turned out to be the expensive part: on the 2026-08-27
-    // v3.8.50 publish `npm pack` alone took 24m37s, leaving 5 of the step's 30-minute
-    // budget for two installs and two boots. Without a duration here the log showed
-    // only "packing…" then a timeout, which reads like a hang and is not.
-    const packStarted = Date.now();
    log(`packing v${version}…`);
    const packOut = execFileSync("npm", ["pack", "--json", "--pack-destination", tmp], {
      cwd: ROOT,
@@ -379,31 +230,13 @@ async function main() {
      maxBuffer: 128 * 1024 * 1024,
    });
    const tarball = path.join(tmp, pickTarball(packOut));
-    const packMb = (fs.statSync(tarball).size / 1024 / 1024).toFixed(1);
-    log(`packed in ${Math.round((Date.now() - packStarted) / 1000)}s (${packMb} MB)`);
-
-    // Each install tree is ~3 GB and this run builds two of them, side by side, plus the
-    // ~275 MB tarball. On the v3.8.50 publish run that overflowed the runner disk mid-way
-    // through the Phase B upgrade install; npm warned per truncated tar entry and still
-    // exited 0, and every later measurement was taken from a broken tree.
-    const availableBytes = freeBytes(tmp);
-    if (availableBytes !== null) {
-      log(`free space in ${tmp}: ${(availableBytes / GB).toFixed(1)} GB`);
-      if (availableBytes < 12 * GB) {
-        warn(
-          `only ${(availableBytes / GB).toFixed(1)} GB free — this gate needs roughly 12 GB ` +
-            `(two ~3 GB install trees, the second installed over twice, plus the tarball). ` +
-            `An install truncated by ENOSPC looks like a schema divergence.`
-        );
-      }
-    }

    // ---- Phase A: clean install -------------------------------------------------
    log("PHASE A — clean install of the packed tarball");
    const aPrefix = path.join(tmp, "a-prefix");
    const aData = path.join(tmp, "a-data");
    fs.mkdirSync(aData, { recursive: true });
-    npmInstallInto(aPrefix, tarball, "clean install");
+    npmInstallInto(aPrefix, tarball);
    const a = await bootAndProbe({
      prefix: aPrefix,
      dataDir: aData,
@@ -412,34 +245,14 @@ async function main() {
      label: "clean",
    });
    failures.push(...a.failures);
-    const cleanBooted = a.ok;
    if (a.ok) log(`clean install healthy on v${a.version}`);
-    else if (a.tail?.length) {
-      console.error("[install-upgrade] last output from the clean-install server:");
-      console.error(a.tail.join("").split("\n").slice(-40).join("\n"));
-    }
    const aDb = findDb(aData);
    const freshTables = aDb ? readTables(aDb) : null;
    if (!freshTables) failures.push("clean: no SQLite database was created");
    else log(`clean install schema: ${freshTables.size} tables`);

-    // Phase A is fully measured (boot verdict + schema snapshot); its ~3 GB install tree is
-    // dead weight from here on and Phase B needs the room. The DATA_DIR stays — only the
-    // node_modules tree goes.
-    if (!skipUpgrade) {
-      fs.rmSync(aPrefix, { recursive: true, force: true });
-      const reclaimed = freeBytes(tmp);
-      log(
-        "released the clean-install tree before the upgrade phase" +
-          (reclaimed !== null ? ` (${(reclaimed / GB).toFixed(1)} GB free)` : "")
-      );
-    }
-
    // ---- Phase B: upgrade over the previous published version -------------------
    let upgradedTables = null;
-    // `--skip-upgrade` never reaches the convergence block (upgradedTables stays null), so
-    // defaulting this to true keeps that path unchanged.
-    let upgradeBooted = true;
    if (skipUpgrade) {
      warn("PHASE B skipped (--skip-upgrade)");
    } else {
@@ -449,7 +262,7 @@ async function main() {
      const bData = path.join(tmp, "b-data");
      fs.mkdirSync(bData, { recursive: true });

-      npmInstallInto(bPrefix, `omniroute@${previous}`, `previous(${previous}) install`);
+      npmInstallInto(bPrefix, `omniroute@${previous}`);
      const before = await bootAndProbe({
        prefix: bPrefix,
        dataDir: bData,
@@ -460,16 +273,14 @@ async function main() {
      if (!before.ok) {
        // A broken PREVIOUS version is not this release's fault — degrade to a warning so a
        // historically bad publish cannot block the current one.
-        warnings.push(
-          `previous version ${previous} did not boot cleanly — upgrade path unverified`
-        );
+        warnings.push(`previous version ${previous} did not boot cleanly — upgrade path unverified`);
        for (const f of before.failures) warn(f);
      } else {
        const beforeDb = findDb(bData);
        const beforeTables = beforeDb ? readTables(beforeDb) : new Set();
        log(`previous(${previous}) schema: ${beforeTables.size} tables — upgrading in place`);

-        npmInstallInto(bPrefix, tarball, "upgrade install");
+        npmInstallInto(bPrefix, tarball);
        const after = await bootAndProbe({
          prefix: bPrefix,
          dataDir: bData,
@@ -479,11 +290,6 @@ async function main() {
        });
        failures.push(...after.failures);
        if (after.ok) log(`upgrade healthy on v${after.version}`);
-        upgradeBooted = after.ok;
-        if (!after.ok && after.tail?.length) {
-          console.error("[install-upgrade] last output from the upgraded server:");
-          console.error(after.tail.join("").split("\n").slice(-40).join("\n"));
-        }

        const afterDb = findDb(bData);
        upgradedTables = afterDb ? readTables(afterDb) : null;
@@ -500,19 +306,7 @@ async function main() {
    }

    // ---- Schema convergence -----------------------------------------------------
-    // Only meaningful when BOTH servers actually served. A boot that died before serving
-    // never ran a migration, so its database still holds the PREVIOUS release's schema and
-    // every post-baseline table shows up as "a clean install creates but an upgrade does
-    // not" — which is what the v3.8.50 publish run reported after ENOSPC truncated the
-    // upgrade install. Comparing there does not add information, it manufactures a
-    // 15-table false alarm on top of the real failure. The run still fails: the boot
-    // failure is already in `failures`.
-    if (freshTables && upgradedTables && !(cleanBooted && upgradeBooted)) {
-      warn(
-        "schema convergence NOT evaluated — a phase failed to boot, so its database was " +
-          "never migrated and any table difference would describe the broken boot, not the schema"
-      );
-    } else if (freshTables && upgradedTables) {
+    if (freshTables && upgradedTables) {
      const verdict = evaluateConvergence({
        freshTables,
        upgradedTables,
@@ -540,10 +334,7 @@ async function main() {

 // Only run the (expensive) gate when invoked directly — importing this module for the pure
 // helper above must not pack, install or boot anything.
-if (
-  process.argv[1] &&
-  path.resolve(process.argv[1]) === path.resolve(new URL(import.meta.url).pathname)
-) {
+if (process.argv[1] && path.resolve(process.argv[1]) === path.resolve(new URL(import.meta.url).pathname)) {
  main().catch((err) => {
    console.error(`[install-upgrade] crashed: ${err?.message ?? err}`);
    process.exit(1);
--- a/scripts/check/check-workflows.mjs
+++ b/scripts/check/check-workflows.mjs
@@ -42,7 +42,6 @@ import { execFileSync, spawnSync } from "node:child_process";
 import fs from "node:fs";
 import path from "node:path";
 import { pathToFileURL } from "node:url";
-import { findProvenanceOnSelfHosted, formatProvenanceFinding } from "./lib/provenanceRunner.mjs";

 const ROOT = process.cwd();
 const WORKFLOWS_DIR = path.join(ROOT, ".github", "workflows");
@@ -276,23 +275,6 @@ export function runZizmor(workflowsDir) {
 // Main
 // ---------------------------------------------------------------------------

-/**
- * Hard rule (not a lint count): `--provenance` inside a job that runs on a
- * self-hosted runner. npm answers 422 at the registry, and in v3.8.50 that
- * answer only came after the tag, the GitHub Release and the Docker images were
- * already out. Blocks under --strict AND --ratchet (the CI mode); plain mode
- * reports it like everything else.
- * @param {string[]} files absolute workflow paths
- */
-export function runProvenanceRunnerCheck(files) {
-  const findings = [];
-  for (const file of files) {
-    const text = fs.readFileSync(file, "utf8");
-    findings.push(...findProvenanceOnSelfHosted(text, path.relative(ROOT, file)));
-  }
-  return findings;
-}
-
 function main() {
  const hasActionlint = isBinaryAvailable("actionlint");
  const hasZizmor = isBinaryAvailable("zizmor");
@@ -368,16 +350,6 @@ function main() {
    }
  }

-  const provenanceFindings = runProvenanceRunnerCheck(workflowFiles);
-  if (provenanceFindings.length > 0) {
-    console.error(
-      `[check-workflows] provenance×self-hosted: ${provenanceFindings.length} finding(s) — HARD RULE:`
-    );
-    provenanceFindings.forEach((f) => console.error(`  ${formatProvenanceFinding(f)}`));
-  } else if (!QUIET) {
-    console.log("[check-workflows] provenance×self-hosted: OK (0 findings)");
-  }
-
  const total = actionlintCount + zizmorCount;
  process.stdout.write(`workflowFindings=${total}\n`);
  process.stdout.write(`actionlintFindings=${actionlintCount}\n`);
@@ -385,15 +357,6 @@ function main() {
  // Read this line with the count above: a finding total is only reproducible against the
  // version that produced it. See zizmorVersion().
  process.stdout.write(`zizmorVersion=${hasZizmor ? zizmorVersion() : "absent"}\n`);
-  process.stdout.write(`provenanceRunnerFindings=${provenanceFindings.length}\n`);
-  if ((STRICT || RATCHET) && provenanceFindings.length > 0) {
-    console.error(
-      `\n[check-workflows] FAIL — ${provenanceFindings.length} job(s) publish with --provenance from a self-hosted runner.\n` +
-        "  npm rejects that with 422 at the registry. Move the upload step to a github-hosted job\n" +
-        "  (see .github/workflows/npm-publish.yml `stage-npm` for the pattern)."
-    );
-    process.exit(1);
-  }

  if (STRICT && total > 0) {
    console.error(`\n[check-workflows] FAIL — ${total} workflow finding(s) total (--strict mode).`);
--- a/scripts/check/lib/configExpiry.mjs
+++ b/scripts/check/lib/configExpiry.mjs
@@ -1,90 +0,0 @@
-/**
- * scripts/check/lib/configExpiry.mjs
- *
- * Finds dated validity fields in JSON config packs so a test can fail BEFORE
- * they lapse. Origin: config/alibaba-free-tier-allowlist.json carried
- * `"validUntil": "2026-08-27"`; on 2026-08-28 the loader started (correctly)
- * rejecting the pack and a test that asserted "the shipped pack loads" turned
- * every PR and main red with no commit involved (#11866). A time bomb, not a
- * regression — and the only kind of defect a diff review can never catch.
- *
- * Pure helpers; the repo-wide assertion lives in
- * tests/unit/config-expiry-time-bomb.test.ts.
- */
-import fs from "node:fs";
-import path from "node:path";
-
-export const EXPIRY_KEY =
-  /^(validUntil|valid_until|validTo|valid_to|expiresAt|expires_at|expiry|expires)$/;
-const DAY_MS = 86_400_000;
-
-/**
- * Walks a parsed JSON value and returns every string-valued expiry field.
- * @returns {{ file: string, keyPath: string, raw: string, expiresAt: number|null }[]}
- */
-export function collectExpiryFields(value, file, keyPath = []) {
-  const out = [];
-  if (Array.isArray(value)) {
-    value.forEach((v, i) => out.push(...collectExpiryFields(v, file, [...keyPath, String(i)])));
-    return out;
-  }
-  if (!value || typeof value !== "object") return out;
-  for (const [key, v] of Object.entries(value)) {
-    const kp = [...keyPath, key];
-    if (EXPIRY_KEY.test(key) && typeof v === "string") {
-      const ms = Date.parse(v);
-      out.push({ file, keyPath: kp.join("."), raw: v, expiresAt: Number.isFinite(ms) ? ms : null });
-    } else if (v && typeof v === "object") {
-      out.push(...collectExpiryFields(v, file, kp));
-    }
-  }
-  return out;
-}
-
-/** @returns {"expired"|"expiring"|"ok"|"unparseable"} */
-export function classifyExpiry(field, nowMs, warnDays = 7) {
-  if (field.expiresAt === null) return "unparseable";
-  if (field.expiresAt < nowMs) return "expired";
-  if (field.expiresAt < nowMs + warnDays * DAY_MS) return "expiring";
-  return "ok";
-}
-
-/** All *.json under dir, recursively, skipping node_modules. Sorted for stable output. */
-export function walkJsonFiles(dir) {
-  const out = [];
-  const stack = [dir];
-  while (stack.length > 0) {
-    const current = stack.pop();
-    let entries;
-    try {
-      entries = fs.readdirSync(current, { withFileTypes: true });
-    } catch {
-      continue;
-    }
-    for (const e of entries) {
-      const full = path.join(current, e.name);
-      if (e.isDirectory()) {
-        if (e.name !== "node_modules") stack.push(full);
-      } else if (e.isFile() && e.name.endsWith(".json")) {
-        out.push(full);
-      }
-    }
-  }
-  return out.sort();
-}
-
-/**
- * Scans every JSON file under `dir`; `file` in the result is relative to `dir`
- * with forward slashes, so allowlists can key on it portably.
- */
-export function scanConfigExpiry(dir) {
-  return walkJsonFiles(dir).flatMap((f) => {
-    let parsed;
-    try {
-      parsed = JSON.parse(fs.readFileSync(f, "utf8"));
-    } catch {
-      return []; // not this scanner's job to validate JSON
-    }
-    return collectExpiryFields(parsed, path.relative(dir, f).split(path.sep).join("/"));
-  });
-}
--- a/scripts/check/lib/provenanceRunner.mjs
+++ b/scripts/check/lib/provenanceRunner.mjs
@@ -1,83 +0,0 @@
-/**
- * scripts/check/lib/provenanceRunner.mjs
- *
- * npm refuses `--provenance` from a self-hosted runner:
- *
- *   422 Unprocessable Entity - Error verifying sigstore provenance bundle:
- *   Unsupported GitHub Actions runner environment: "self-hosted".
- *   Only "github-hosted" runners are supported when publishing with provenance.
- *
- * v3.8.50 hit this at the very end of a 76-minute publish job — after the tag,
- * the GitHub Release and the Docker images were already public — because
- * `USE_VPS_RUNNER` had been turned on (2026-08-02) with no release in between to
- * surface it. The combination is greppable, so it must fail in CI the moment a
- * workflow introduces it, not four weeks later at the registry.
- *
- * Pure: takes workflow YAML text, returns the offending (job, step) pairs.
- */
-import { load as yamlLoad } from "js-yaml";
-
-const SELF_HOSTED = /\bself-hosted\b/;
-const EXPRESSION = /\$\{\{/;
-// Lookahead, not \b: `--provenance-file=…` is a different flag (a pre-built
-// bundle) and must not match — a word boundary sits between "e" and "-".
-const PROVENANCE = /(^|\s)--provenance(?=\s|=|$)/m;
-
-/**
- * Classifies a job's `runs-on` value.
- * @returns {"self-hosted"|"hosted"|"unknown"}
- *   "unknown" = an expression with no literal `self-hosted` in it (e.g.
- *   `${{ matrix.os }}`); the check does not guess, it skips.
- */
-export function classifyRunsOn(runsOn) {
-  if (runsOn == null) return "unknown";
-  if (typeof runsOn === "string") {
-    if (SELF_HOSTED.test(runsOn)) return "self-hosted";
-    return EXPRESSION.test(runsOn) ? "unknown" : "hosted";
-  }
-  if (Array.isArray(runsOn)) {
-    return runsOn.some((v) => typeof v === "string" && SELF_HOSTED.test(v))
-      ? "self-hosted"
-      : "hosted";
-  }
-  if (typeof runsOn === "object") {
-    // { group: ..., labels: ... } form
-    const labels = runsOn.labels;
-    return classifyRunsOn(Array.isArray(labels) ? labels : labels == null ? "" : String(labels));
-  }
-  return "unknown";
-}
-
-/**
- * @param {string} yamlText
- * @param {string} fileName  used only for reporting
- * @returns {{ file: string, job: string, step: string }[]}
- */
-export function findProvenanceOnSelfHosted(yamlText, fileName = "<workflow>") {
-  let doc;
-  try {
-    doc = yamlLoad(yamlText);
-  } catch {
-    // actionlint owns syntax; an unparseable file is not this rule's finding.
-    return [];
-  }
-  const jobs =
-    doc && typeof doc === "object" && doc.jobs && typeof doc.jobs === "object" ? doc.jobs : {};
-  const findings = [];
-  for (const [jobName, job] of Object.entries(jobs)) {
-    if (!job || typeof job !== "object") continue;
-    if (classifyRunsOn(job["runs-on"]) !== "self-hosted") continue;
-    const steps = Array.isArray(job.steps) ? job.steps : [];
-    steps.forEach((step, i) => {
-      if (step && typeof step.run === "string" && PROVENANCE.test(step.run)) {
-        findings.push({ file: fileName, job: jobName, step: step.name || `#${i + 1}` });
-      }
-    });
-  }
-  return findings;
-}
-
-/** Human-readable line per finding, used by the CLI. */
-export function formatProvenanceFinding(f) {
-  return `${f.file}: job "${f.job}", step "${f.step}" runs \`--provenance\` on a self-hosted runner — npm rejects that (422). Move the upload to a github-hosted job.`;
-}
--- a/scripts/dev/smoke-electron-packaged.mjs
+++ b/scripts/dev/smoke-electron-packaged.mjs
@@ -123,48 +123,6 @@ function discoverPackagedExecutable() {
  throw new Error(`Packaged Electron smoke check does not support ${platform()}.`);
 }

-/**
- * The packaged app opens SQLite lazily: `/login` (the readiness URL) never touches the
- * database, so a smoke that only waits for readiness sees no `[DB]` line at all. After
- * readiness the smoke requests a DB-backed endpoint and waits for evidence that the
- * database opened. The primary open path does NOT print "[DB] Driver: ..." (only the
- * recovery path and the sql.js fallback do), so the evidence is any `[DB]`/`[Migration]`
- * startup line — and the #7592 guard below rejects the fallback's own line explicitly.
- */
-export const DB_TOUCH_PATH = "/api/monitoring/health";
-export const DB_OPEN_EVIDENCE_PATTERN =
-  /\[DB\] (Driver: |SQLite database ready|Added [^\n]* column|Changing cache_size|cache_size changed)|\[Migration\] (Applied|Pre-migration backup)/;
-
-export async function waitForDatabaseOpen(getLogs, { timeoutMs = 15_000, pollMs = 250 } = {}) {
-  const startedAt = Date.now();
-  while (Date.now() - startedAt < timeoutMs) {
-    const logs = getLogs();
-    assertNoFatalLogs(logs);
-    if (DB_OPEN_EVIDENCE_PATTERN.test(logs)) return logs;
-    await sleep(pollMs);
-  }
-  throw new Error(
-    `Packaged Electron app logged no [DB]/[Migration] startup line within ${timeoutMs}ms of ` +
-      `touching ${DB_TOUCH_PATH} — the database never opened, so the SQLite driver cannot be verified.`
-  );
-}
-
-async function openDatabaseForSmoke({ logs, smokeUrl }) {
-  const touchUrl = new URL(DB_TOUCH_PATH, smokeUrl).toString();
-  try {
-    const response = await fetchWithTimeout(touchUrl, 5_000);
-    console.log(
-      `[electron-smoke] touched ${touchUrl} (HTTP ${response.status}) to open the database`
-    );
-  } catch (error) {
-    console.log(
-      `[electron-smoke] touching ${touchUrl} failed (${error instanceof Error ? error.message : String(error)}) — waiting for the database anyway`
-    );
-  }
-  await waitForDatabaseOpen(() => logs.value);
-  console.log("[electron-smoke] database opened");
-}
-
 async function fetchWithTimeout(url, timeoutMs) {
  const controller = new AbortController();
  const timeout = setTimeout(() => controller.abort(), timeoutMs);
@@ -492,13 +450,8 @@ export function assertNativeDriverSelected(logs) {
    );
  }

-  // The primary open path prints no "[DB] Driver: ..." line at all (only the recovery path and
-  // the sql.js fallback do), so a database that demonstrably opened WITHOUT the fallback's own
-  // line is the native driver — that is exactly what #7592 guards.
-  if (DB_OPEN_EVIDENCE_PATTERN.test(logs)) return;
-
  throw new Error(
-    "Packaged Electron app logs show no database activity at all — cannot confirm which SQLite " +
+    "Packaged Electron app logs contain no '[DB] Driver: ...' line — cannot confirm which SQLite " +
      "driver loaded."
  );
 }
@@ -553,14 +506,7 @@ async function waitForReady({ logs, smokeUrl, timeoutMs, settleMs, exitState })
 * by the single-launch path and the cold-restart (two-launch) path so both
 * exercise identical spawn/readiness/shutdown behavior.
 */
-async function launchAndCollectLogs({
-  appExecutable,
-  smokeUrl,
-  dataDir,
-  timeoutMs,
-  settleMs,
-  streamLogs,
-}) {
+async function launchAndCollectLogs({ appExecutable, smokeUrl, dataDir, timeoutMs, settleMs, streamLogs }) {
  const smokeEnv = buildSmokeEnv({ dataDir });
  await assertPortIsFree(smokeUrl);
  await ensureSmokeEnvDirs(smokeEnv, dataDir);
@@ -592,8 +538,6 @@ async function launchAndCollectLogs({

  try {
    await waitForReady({ logs, smokeUrl, timeoutMs, settleMs, exitState });
-    // Outside waitForReady on purpose: a missing database is a verdict, not a readiness retry.
-    await openDatabaseForSmoke({ logs, smokeUrl });
    return logs.value;
  } catch (error) {
    if (!streamLogs) {
@@ -624,14 +568,7 @@ async function main() {
    !process.env.ELECTRON_SMOKE_DATA_DIR && process.env.ELECTRON_SMOKE_KEEP_DATA !== "1";

  try {
-    await launchAndCollectLogs({
-      appExecutable,
-      smokeUrl,
-      dataDir,
-      timeoutMs,
-      settleMs,
-      streamLogs,
-    });
+    await launchAndCollectLogs({ appExecutable, smokeUrl, dataDir, timeoutMs, settleMs, streamLogs });

    if (!coldRestart) return;

--- a/scripts/ops/runner-janitor.sh
+++ b/scripts/ops/runner-janitor.sh
@@ -1,172 +1,53 @@
 #!/usr/bin/env bash
-# runner-janitor — self-hosted runner box hygiene for the .113 pool.
+# runner-janitor — self-hosted runner box hygiene (WS3.3, v3.8.49 quality plan).
 #
-# Runs from cron every 30 min (see docs/ops/RUNNER_BOX.md). It ACTS on what it
-# can prove is safe and ALERTS on what needs an operator decision. Reads of
-# "is this in use?" and the removal happen in the same command, never in two
-# passes: a check-then-delete with a gap is how a live Build job lost its _work
-# on 2026-08-27.
+# The .113 runner box has recurring failure modes that until now were manual
+# discipline: orphaned tmpfs/work dirs filling the disk, and >4 concurrent
+# runners OOM-killing jobs (16 GB box; incidents on the v3.8.47 release day).
+# Install via cron on the box (see docs/ops/RUNNER_BOX.md):
+#   */30 * * * *  /opt/omniroute-ops/runner-janitor.sh >> /var/log/runner-janitor.log 2>&1
 #
-# Measured box (2026-08-28): 31 GB RAM, 32 cores, 15 GB swap, /tmp = 12 GB
-# tmpfs (RAM!), 188 GB disk. A single `next-build` peaks at ~14 GB, so two
-# concurrent heavy builds saturate the box and three take it down (06:42Z that
-# day: load 56, two jobs lost). The v3.8.50 postmortem (Parte III) has the numbers.
-#
-# What it does, in order:
-#   1) sweep stale artefacts our tooling leaves behind — tmpfs bases after 3 h
-#      (they hold RAM), disk _work/_temp bases after 24 h; only names we create,
-#      only when no process has them open
-#   2) kill zombie builds: a `next-build` older than ZOMBIE_BUILD_MAX_MIN has no
-#      job attached (a real Build step measures ~26 min). On 2026-08-27 one ran
-#      70 minutes after GitHub had already declared its job lost, eating 3.6 GB
-#      and a full core set. KillMode=mixed on the units covers systemctl
-#      stop/restart; this covers the lost-connection path.
-#   3) prune 48 h-old checkouts under _work of runners whose unit is INACTIVE
-#      (stopped runners cannot be mid-job; active ones are never touched)
-#   4) alert: root disk >= DISK_ALERT_PCT, memory PSI full/avg60 >= threshold,
-#      Runner.Listener count above the ceiling (with a per-project breakdown —
-#      the box also hosts OmniHeuris and OmniMind runners)
-#
-# Usage: runner-janitor.sh [--dry-run] [--help]
 # Exit codes: 0 healthy · 1 attention needed (printed to stdout for the log).
 set -euo pipefail

-DRY_RUN=0
-for arg in "$@"; do
-  case "$arg" in
-    --dry-run) DRY_RUN=1 ;;
-    -h|--help)
-      sed -n '2,32p' "$0" | sed 's/^# \{0,1\}//'
-      exit 0 ;;
-    *) echo "unknown argument: $arg" >&2; exit 2 ;;
-  esac
-done
-
-MAX_ACTIVE_RUNNERS="${MAX_ACTIVE_RUNNERS:-8}"
+MAX_ACTIVE_RUNNERS="${MAX_ACTIVE_RUNNERS:-4}"
 DISK_ALERT_PCT="${DISK_ALERT_PCT:-85}"
-TMPFS_MAX_AGE_HOURS="${TMPFS_MAX_AGE_HOURS:-3}"
-WORK_TEMP_MAX_AGE_HOURS="${WORK_TEMP_MAX_AGE_HOURS:-24}"
-WORK_CHECKOUT_MAX_AGE_HOURS="${WORK_CHECKOUT_MAX_AGE_HOURS:-48}"
-ZOMBIE_BUILD_MAX_MIN="${ZOMBIE_BUILD_MAX_MIN:-75}"
-ZOMBIE_BUILD_COMM="${ZOMBIE_BUILD_COMM:-next-build}"
-PSI_FULL_AVG60_ALERT="${PSI_FULL_AVG60_ALERT:-10}"
-# Overridable so the unit test can point everything at a fixture tree.
-JANITOR_TMP_BASES="${JANITOR_TMP_BASES-/tmp}"
-JANITOR_WORK_TEMP_BASES="${JANITOR_WORK_TEMP_BASES-/opt/actions-runner*/_work/_temp /home/*/actions-runner*/_work/_temp}"
-JANITOR_RUNNER_DIRS="${JANITOR_RUNNER_DIRS-/opt/actions-runner*}"
-JANITOR_PSI_FILE="${JANITOR_PSI_FILE:-/proc/pressure/memory}"
-JANITOR_DF_PATH="${JANITOR_DF_PATH:-/}"
-
+WORK_DIR_MAX_AGE_HOURS="${WORK_DIR_MAX_AGE_HOURS:-24}"
 STATUS=0
-say() { echo "[janitor] $*"; }

-# "Is anything using this?" — ONE snapshot of every open path on the box
-# (lsof -Fn), then a prefix match per candidate. `lsof +D <dir>` walks the whole
-# tree instead and took minutes on a 5 GB leftover — unusable from cron. An
-# absent lsof means "cannot prove idle": the sweep keeps the path and says so.
-LSOF_BIN="${JANITOR_LSOF:-lsof}"
-have_busy_tools() { command -v "$LSOF_BIN" >/dev/null 2>&1; }
-SNAP=""
-cleanup() { [ -n "$SNAP" ] && rm -f -- "$SNAP"; }
-trap cleanup EXIT
-# One lsof for the whole run (~13 s / 83k lines on the box), kept ONLY for the
-# bases we sweep — 460 candidates grepping a re-printed 83k-line string was the
-# slow part, not lsof itself.
-snapshot_open_paths() {
-  have_busy_tools || return 0
-  SNAP=$(mktemp) || return 0
-  local prefixes="" b
-  for b in $JANITOR_TMP_BASES $JANITOR_WORK_TEMP_BASES; do [ -d "$b" ] && prefixes="$prefixes"$'\n'"$b/"; done
-  # -F n: one "n<path>" line per open file; -w: no warnings
-  "$LSOF_BIN" -w -Fn 2>/dev/null | sed -n 's/^n//p' | grep -F -f <(printf '%s' "$prefixes" | sed '/^$/d') > "$SNAP" 2>/dev/null || true
-}
-is_busy() {
-  local p="$1"
-  [ -n "$SNAP" ] && [ -s "$SNAP" ] || return 1
-  # exact path, or anything beneath it when it is a directory
-  grep -qxF -- "$p" "$SNAP" && return 0
-  [ -d "$p" ] && grep -qF -- "$p/" "$SNAP"
-}
+echo "[janitor] $(date -u +%FT%TZ) start"

-# sweep <base> <max-age-minutes>: only names our tooling creates, never through
-# a symlinked base, never across a filesystem, and remove+check in one step.
-sweep() {
-  local base="$1" max_min="$2" p
-  [ -d "$base" ] || return 0
-  [ -L "$base" ] && { say "skip symlinked base: $base"; return 0; }
-  while IFS= read -r -d '' p; do
-    if ! have_busy_tools; then say "cannot prove idle (lsof missing — apt install lsof), kept: $p"; STATUS=1; continue; fi
-    if is_busy "$p"; then say "busy, kept: $p"; continue; fi
-    if [ "$DRY_RUN" -eq 1 ]; then say "would remove ($(( max_min / 60 ))h+): $p"; else rm -rf -- "$p" && say "removed ($(( max_min / 60 ))h+): $p"; fi
-  done < <(find -P "$base" -xdev -mindepth 1 -maxdepth 1 \
-      \( -name 'runner-*' -o -name 'omniroute-*' -o -name 'next-build*' -o -name 'e2e-build.tar.gz' \) \
-      ! -type l -mmin "+$max_min" -print0 2>/dev/null || true)
-}
-
-say "$(date -u +%FT%TZ) start${DRY_RUN:+ (dry-run=$DRY_RUN)} busy-tools=$(have_busy_tools && echo ok || echo MISSING)"
-
-# 1) stale artefacts — tmpfs is RAM, so it gets the short fuse
-snapshot_open_paths
-for base in $JANITOR_TMP_BASES; do sweep "$base" $(( TMPFS_MAX_AGE_HOURS * 60 )); done
-for base in $JANITOR_WORK_TEMP_BASES; do sweep "$base" $(( WORK_TEMP_MAX_AGE_HOURS * 60 )); done
-say "stale temp sweep done"
-
-# 2) zombie builds
-ZOMBIES=0
-while read -r pid etimes comm; do
-  [ -n "${pid:-}" ] || continue
-  if [ "$etimes" -gt $(( ZOMBIE_BUILD_MAX_MIN * 60 )) ]; then
-    say "⚠ zombie build pid=$pid comm=$comm age=$(( etimes / 60 ))min > ${ZOMBIE_BUILD_MAX_MIN}min — no job runs this long"
-    if [ "$DRY_RUN" -eq 1 ]; then say "[dry-run] would: kill -TERM $pid (then -KILL)"; else
-      kill -TERM "$pid" 2>/dev/null || true; sleep 10
-      kill -0 "$pid" 2>/dev/null && { kill -KILL "$pid" 2>/dev/null || true; say "  needed SIGKILL"; }
-    fi
-    ZOMBIES=$(( ZOMBIES + 1 )); STATUS=1
-  fi
-done < <(ps -eo pid=,etimes=,comm= 2>/dev/null | awk -v c="$ZOMBIE_BUILD_COMM" '$3 ~ ("^" c) {print $1, $2, $3}' || true)
-say "zombie builds: $ZOMBIES"
-
-# 3) old checkouts of STOPPED runners
-for d in $JANITOR_RUNNER_DIRS; do
-  [ -d "$d" ] && [ -f "$d/.runner" ] || continue
-  agent=$(grep -o '"agentName": *"[^"]*"' "$d/.runner" 2>/dev/null | sed 's/.*"\([^"]*\)"$/\1/')
-  [ -n "$agent" ] || continue
-  unit=$(systemctl list-units --plain --no-legend "actions.runner.*.${agent}.service" 2>/dev/null | awk 'NR==1{print $1}')
-  [ -n "$unit" ] || continue
-  if systemctl is-active --quiet "$unit"; then continue; fi
-  while IFS= read -r -d '' co; do
-    if [ "$DRY_RUN" -eq 1 ]; then say "would prune checkout of stopped runner $agent: $co"; else rm -rf -- "$co" && say "pruned checkout of stopped runner $agent: $co"; fi
-  done < <(find -P "$d/_work" -xdev -mindepth 2 -maxdepth 2 -type d -mmin "+$(( WORK_CHECKOUT_MAX_AGE_HOURS * 60 ))" -print0 2>/dev/null || true)
+# 1) Sweep stale runner temp/work leftovers (>24h — no legitimate job runs that long).
+# Hardened for a root cron on world-writable paths: never follow a symlinked base
+# (a compromised runner could plant one), -P + -xdev so the sweep cannot traverse
+# out of the filesystem, and patterns narrowed to names OUR tooling creates
+# (no generic tmp* — unrelated system temp files are out of scope).
+for base in /tmp /home/*/actions-runner*/_work/_temp; do
+  [ -d "$base" ] || continue
+  [ -L "$base" ] && { echo "[janitor] skip symlinked base: $base"; continue; }
+  find -P "$base" -xdev -maxdepth 1 \( -name 'runner-*' -o -name 'omniroute-*' \) \
+    ! -type l -mmin +$((WORK_DIR_MAX_AGE_HOURS * 60)) -exec rm -rf {} + 2>/dev/null || true
 done
+echo "[janitor] stale temp sweep done"

-# 4a) disk
-USAGE=$(df --output=pcent "$JANITOR_DF_PATH" 2>/dev/null | tail -1 | tr -dc '0-9')
-if [ "${USAGE:-0}" -ge "$DISK_ALERT_PCT" ]; then
-  say "⚠ ROOT DISK ${USAGE}% >= ${DISK_ALERT_PCT}% — clean before the next heavy run"; STATUS=1
-else
-  say "disk ${USAGE:-?}% OK"
-fi
-
-# 4b) memory pressure (PSI) — the box swapped its way through the v3.8.50 publish
-if [ -r "$JANITOR_PSI_FILE" ]; then
-  FULL60=$(awk '/^full/ {for(i=1;i<=NF;i++) if ($i ~ /^avg60=/) {sub("avg60=","",$i); print $i}}' "$JANITOR_PSI_FILE" 2>/dev/null || echo "")
-  if [ -n "$FULL60" ] && awk -v v="$FULL60" -v t="$PSI_FULL_AVG60_ALERT" 'BEGIN{exit !(v+0 >= t+0)}'; then
-    say "⚠ MEMORY PRESSURE psi full/avg60=${FULL60}% >= ${PSI_FULL_AVG60_ALERT}% — too many heavy jobs at once"; STATUS=1
-  else
-    say "memory psi full/avg60=${FULL60:-n/a}% OK"
-  fi
-fi
-
-# 4c) concurrency ceiling — alert with a breakdown; the fix is fewer/labelled
-#     runners (an operator decision), not killing listeners from cron.
-ACTIVE=$(pgrep -fc "Runner.Listener" || true)
-OMNI=$(pgrep -fc "actions-runner-omniroute[^ ]*/bin[^ ]*/Runner.Listener" || true)
-if [ "${ACTIVE:-0}" -gt "$MAX_ACTIVE_RUNNERS" ]; then
-  say "⚠ ${ACTIVE} Runner.Listener processes (omniroute=${OMNI:-0}, other=$(( ${ACTIVE:-0} - ${OMNI:-0} ))) > ceiling ${MAX_ACTIVE_RUNNERS} — stop idle extras: systemctl stop <unit> only when it has no Runner.Worker child"
+# 2) Disk pressure — alert loudly before SQLITE_FULL kills jobs mid-run.
+USAGE=$(df --output=pcent / | tail -1 | tr -dc '0-9')
+if [ "$USAGE" -ge "$DISK_ALERT_PCT" ]; then
+  echo "[janitor] ⚠ ROOT DISK ${USAGE}% >= ${DISK_ALERT_PCT}% — clean before the next heavy run"
  STATUS=1
 else
-  say "runners active: ${ACTIVE:-0}/${MAX_ACTIVE_RUNNERS} (omniroute=${OMNI:-0}) OK"
+  echo "[janitor] disk ${USAGE}% OK"
 fi

-say "done status=$STATUS"
+# 3) Concurrency ceiling — 8-wide OOMed the 16 GB box twice on release day;
+#    4 is the proven ceiling. This CODIFIES the rule that was manual discipline.
+ACTIVE=$(pgrep -fc "Runner.Listener" || true)
+if [ "${ACTIVE:-0}" -gt "$MAX_ACTIVE_RUNNERS" ]; then
+  echo "[janitor] ⚠ ${ACTIVE} Runner.Listener processes > ceiling ${MAX_ACTIVE_RUNNERS} — stop the extra runners (systemctl stop actions.runner.<name>)"
+  STATUS=1
+else
+  echo "[janitor] runners active: ${ACTIVE:-0}/${MAX_ACTIVE_RUNNERS} OK"
+fi
+
+echo "[janitor] done status=$STATUS"
 exit "$STATUS"
--- a/src/lib/db/migrations/163_model_capabilities.sql
+++ b/src/lib/db/migrations/163_model_capabilities.sql
@@ -1,43 +0,0 @@
-- 163_model_capabilities.sql
--
-- Promote `model_capabilities` from a lazily-created runtime table to a real migration.
--
-- WHY: the table was only ever created by `ensureCapabilitiesTable()` in
-- src/lib/modelsDevSync.ts, on demand, the first time a models.dev capability sync ran.
-- Whether a database has it therefore depends on TIMING, not on the schema version — so a
-- clean install and an upgraded install diverge for no structural reason. The v3.8.50
-- publish run hit exactly that: `check:install-upgrade` reported `model_capabilities` as a
-- table "present only after upgrade", because the older database had already run a sync
-- and the freshly-installed one had not.
--
-- Creating it here makes both install paths converge deterministically.
-- `ensureCapabilitiesTable()` stays in place as an idempotent safety net (it is a
-- CREATE TABLE IF NOT EXISTS and now always a no-op); tests/unit/db-install-upgrade-schema-parity.test.ts
-- pins the two definitions against drift.
--
-- IF NOT EXISTS is required, not decorative: every database that ever ran a models.dev
-- sync already has this table, and this migration must be a no-op there.
-
-CREATE TABLE IF NOT EXISTS model_capabilities (
-  provider TEXT NOT NULL,
-  model_id TEXT NOT NULL,
-  tool_call BOOLEAN,
-  reasoning BOOLEAN,
-  attachment BOOLEAN,
-  structured_output BOOLEAN,
-  temperature BOOLEAN,
-  modalities_input TEXT,
-  modalities_output TEXT,
-  knowledge_cutoff TEXT,
-  release_date TEXT,
-  last_updated TEXT,
-  status TEXT,
-  family TEXT,
-  open_weights BOOLEAN,
-  limit_context INTEGER,
-  limit_input INTEGER,
-  limit_output INTEGER,
-  interleaved_field TEXT,
-  last_synced TEXT,
-  PRIMARY KEY (provider, model_id)
-);
--- a/tests/_setup/isolateDataDir.ts
+++ b/tests/_setup/isolateDataDir.ts
@@ -33,7 +33,7 @@ if (!process.env.DATA_DIR) {
  // Best-effort cleanup so a long suite run does not leak hundreds of temp DBs.
  process.on("exit", () => {
    try {
-      fs.rmSync(dir, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+      fs.rmSync(dir, { recursive: true, force: true });
    } catch {
      // ignore — the OS reaps its temp dir eventually.
    }
--- a/tests/e2e/protocol-clients.test.ts
+++ b/tests/e2e/protocol-clients.test.ts
@@ -92,7 +92,7 @@ describe("Protocol clients E2E", () => {
      method: "PATCH",
      body: JSON.stringify({ a2aEnabled: true }),
    });
-    expect([200, 401, 403]).toContain(response.status);
+    expect([200, 401]).toContain(response.status);
  });

  it(
@@ -134,7 +134,7 @@ describe("Protocol clients E2E", () => {
      }

      const auditRes = await apiFetch("/api/mcp/audit?limit=50&tool=omniroute_get_health");
-      expect([200, 401, 403]).toContain(auditRes.status);
+      expect([200, 401]).toContain(auditRes.status);
      if (auditRes.status === 200) {
        expect(auditRes.ok).toBe(true);
        const auditJson = (await auditRes.json()) as any;
--- a/tests/e2e/system-failover.test.ts
+++ b/tests/e2e/system-failover.test.ts
@@ -366,7 +366,7 @@ test.after(async () => {
  await serverA.stop();
  await serverB.stop();
  core.closeDbInstance();
-  await fsp.rm(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+  await fsp.rm(TEST_DATA_DIR, { recursive: true, force: true });
 });

 test("primary healthy: request routes to Server A only", async () => {
--- a/tests/integration/_chatPipelineHarness.ts
+++ b/tests/integration/_chatPipelineHarness.ts
@@ -286,7 +286,7 @@ export async function createChatPipelineHarness(prefix) {
    clearSkillState();
    await new Promise((resolve) => setTimeout(resolve, 20));
    core.resetDbInstance();
-    fs.rmSync(testDataDir, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+    fs.rmSync(testDataDir, { recursive: true, force: true });
    fs.mkdirSync(testDataDir, { recursive: true });
    initTranslators();
  }
@@ -300,7 +300,7 @@ export async function createChatPipelineHarness(prefix) {
    clearSkillState();
    resetAllCircuitBreakers();
    core.resetDbInstance();
-    fs.rmSync(testDataDir, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+    fs.rmSync(testDataDir, { recursive: true, force: true });
  }

  async function seedConnection(provider: string, overrides: SeedConnectionOverrides = {}) {
--- a/tests/integration/agent-bridge-bypass-flow.test.ts
+++ b/tests/integration/agent-bridge-bypass-flow.test.ts
@@ -24,7 +24,7 @@ const DEFAULT_PATTERNS = [".bank.", ".gov.", "okta.com", "auth0.com"];

 function resetDb() {
  core.resetDbInstance();
-  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true });
  fs.mkdirSync(TEST_DATA_DIR, { recursive: true });
 }

@@ -34,7 +34,7 @@ test.beforeEach(() => {
 });

 test.after(() => {
-  try { fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 }); } catch { /* noop */ }
+  try { fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true }); } catch { /* noop */ }
 });

 // ── POST patterns ──────────────────────────────────────────────────────────
--- a/tests/integration/agent-bridge-cert-flow.test.ts
+++ b/tests/integration/agent-bridge-cert-flow.test.ts
@@ -30,7 +30,7 @@ function certFilePath() {
 }

 function resetCertDir() {
-  fs.rmSync(certDir(), { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+  fs.rmSync(certDir(), { recursive: true, force: true });
  fs.mkdirSync(certDir(), { recursive: true });
 }

@@ -39,7 +39,7 @@ test.beforeEach(() => {
 });

 test.after(() => {
-  try { fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 }); } catch { /* noop */ }
+  try { fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true }); } catch { /* noop */ }
 });

 // ── GET /cert ─────────────────────────────────────────────────────────────
--- a/tests/integration/agent-bridge-mappings.test.ts
+++ b/tests/integration/agent-bridge-mappings.test.ts
@@ -24,7 +24,7 @@ const mappingsRoute = await import(

 function resetDb() {
  core.resetDbInstance();
-  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true });
  fs.mkdirSync(TEST_DATA_DIR, { recursive: true });
 }

@@ -33,7 +33,7 @@ test.beforeEach(() => {
 });

 test.after(() => {
-  try { fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 }); } catch { /* noop */ }
+  try { fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true }); } catch { /* noop */ }
 });

 // ── GET (empty) ────────────────────────────────────────────────────────────
--- a/tests/integration/agent-bridge-routes.test.ts
+++ b/tests/integration/agent-bridge-routes.test.ts
@@ -39,7 +39,7 @@ const routeGuard = await import("../../src/server/authz/routeGuard.ts");

 function resetDb() {
  core.resetDbInstance();
-  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true });
  fs.mkdirSync(TEST_DATA_DIR, { recursive: true });
 }

@@ -49,7 +49,7 @@ test.beforeEach(() => {

 test.after(() => {
  try {
-    fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+    fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true });
  } catch {
    /* noop */
  }
--- a/tests/integration/all-statuses-route.test.ts
+++ b/tests/integration/all-statuses-route.test.ts
@@ -39,7 +39,7 @@ async function resetStorage() {
  delete process.env.INITIAL_PASSWORD;
  core.resetDbInstance();
  apiKeysDb.resetApiKeyState();
-  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true });
  fs.mkdirSync(TEST_DATA_DIR, { recursive: true });
 }

@@ -55,7 +55,7 @@ test.beforeEach(async () => {

 test.after(async () => {
  await resetStorage();
-  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true });
 });

 // ── Auth tests ────────────────────────────────────────────────────────────────
@@ -288,6 +288,6 @@ test("grok-build status uses GROK_HOME and returns its managed endpoint", async
  } finally {
    if (original === undefined) delete process.env.GROK_HOME;
    else process.env.GROK_HOME = original;
-    fs.rmSync(grokHome, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
+    fs.rmSync(grokHome, { recursive: true, force: true });
  }
 });
--- a/Show More
+++ b/Show More
				`@@ -1 +0,0 @@`
				- fix(docker): bump the Bun image to 1.4.0, enable Turbopack on Bun, and port the node image's build memory guards so the `-bun` container builds fit the 16 GB GitHub runner instead of dying with `cannot allocate memory` ([#11719](https://github.com/diegosouzapw/OmniRoute/pull/11719)). Both images now default `OMNIROUTE_BUILD_WORKERS` to `2` (1 page-data worker) against the measured ~4.5 GB per-process RSS budget (#7518/#11663).
				`@@ -1 +0,0 @@`
				- Electron release: `electron/package-lock.json` regained the optional `electron-builder-squirrel-windows` subtree (13 entries) that `npm ci` had been refusing as out of sync — the Linux desktop leg died on it — and `electron-release.yml` gained a `build_ref` dispatch input so a release whose tag was cut with the broken lock can have its assets rebuilt from the repaired line
				`@@ -1 +0,0 @@`
				- Electron release workflow: the `publish-npm` job now grants `actions: read` to the reusable `npm-publish.yml` it calls (its `publish` job requests it), which is what made GitHub refuse the whole v3.8.50 run at startup and ship the release with zero desktop assets; a `workflow_dispatch` now builds the requested tag instead of the dispatching branch and can skip the npm leg (`publish_npm=false`) when only re-attaching assets
				`@@ -1 +0,0 @@`
				- npm publish workflow: the CycloneDX SBOM is attached to the GitHub Release on `workflow_dispatch` publishes too (when a release for the tag exists), not only on the `release` event — v3.8.50 shipped through a staged dispatch and its release carried no SBOM until it was attached by hand from the run's `sbom-npm` artifact
				`@@ -1 +0,0 @@`
				- `Coverage` job on `ci.yml`: the informational Codecov upload gets its own 5-minute ceiling and `continue-on-error`, and the job budget grows from 20 to 30 minutes (the 8-shard c8 merge alone takes ~10) — a stalled upload no longer ends the job `cancelled` and drags a fully green `main` run's conclusion down with it