fix(providers): correct opencode-zen muse-spark context length and Responses auth header (#12681, #12633)

- Declare the real ~1M contextLength/maxOutputTokens on the muse-spark-1.2 /
  muse-spark-1.2-contributor-free registry entries (opencode + opencode-zen)
  instead of silently falling back to the 200000 provider default (#12681).
- Send x-api-key instead of Authorization: Bearer for the openai-responses
  format on the main OpenCode Zen host, fixing a 401 on Muse Spark
  Contributor's /v1/responses route; scoped by baseUrl so opencode-go (a
  different upstream) keeps Bearer (#12633).

.env.example

@@ -2721,10 +2721,6 @@ PLAYGROUND_COMPARE_MAX_COLUMNS=4
 # MEMORY_VEC_TOP_K=20                           # default top-K for vector search
 # MEMORY_RRF_K=60                               # RRF k constant (sqlite-vec hybrid recipe)
 # HF_HUB_ENDPOINT=https://huggingface.co        # override Hugging Face Hub base URL for static potion downloads
-# Test/diagnostic seam (src/lib/memory/vectorStore.ts) — forces getVectorStore() to
-# return null (simulates a cloud/WASM environment without sqlite-vec), degrading
-# memory retrieval to FTS5 keyword search. Default off; leave unset in production.
-# VECTOR_STORE_DISABLE_VEC=false
 # TV6 typed memory decay (OPT-IN, default off — the sweep DELETES decayed memories)
 # MEMORY_TYPED_DECAY_ENABLED=false              # master switch for the destructive sweep (default off)
 # MEMORY_TYPED_DECAY_EPISODIC_DAYS=30           # episodic TTL in days; 0 = episodic immune too

changelog.d/fixes/12633-opencode-zen-responses-auth-header.md

@@ -0,0 +1 @@
+- fix(providers): send `x-api-key` instead of `Authorization: Bearer` for OpenCode Zen's `/v1/responses` endpoint (Muse Spark Contributor models), fixing a 401 on OmniRoute's auth header (#12633)

changelog.d/fixes/12681-opencode-muse-spark-context-length.md

@@ -0,0 +1 @@
+- fix(models): declare the real ~1M contextLength for OpenCode Zen's Muse Spark 1.2 models instead of falling back to the 200000 provider default (#12681)

changelog.d/fixes/12745-memory-rerank-loopback-auth.md

@@ -1 +0,0 @@
-- fix(memory): authenticate the internal /v1/rerank loopback call so memory reranking no longer silently degrades to unranked order when REQUIRE_API_KEY=true (#12745)

docs/reference/ENVIRONMENT.md

@@ -923,7 +923,6 @@ Embedding layer, vector store and reranking knobs for the persistent memory subs
 | `HF_HUB_ENDPOINT`               | `https://huggingface.co`   | Override Hugging Face Hub base URL used by `staticPotion.ts` (e.g. mirror endpoint for air-gapped setups). |
 | `MEMORY_VEC_TOP_K`              | `20`                       | Default top-K used by the `sqlite-vec` brute-force vector search inside `src/lib/memory/vectorStore.ts`.   |
 | `MEMORY_RRF_K`                  | `60`                       | Reciprocal Rank Fusion constant `k` for hybrid FTS5 + vector retrieval (sqlite-vec recipe).                |
-| `VECTOR_STORE_DISABLE_VEC`      | `false`                    | Test/diagnostic seam in `getVectorStore()` (`src/lib/memory/vectorStore.ts`): when `true`, forces the vector store to `null` (simulates a cloud/WASM environment without `sqlite-vec`), degrading memory retrieval to FTS5 keyword search. Leave unset in production. |
 | `NOTION_API_KEY`                | _(unset)_                  | API key for Notion backend (used by `genericBackend.ts` known backend preset).                                |
 | `NOTION_API_URL`                | `https://api.notion.com/v1`| Base URL for Notion API (can override for self-hosted Notion alternatives).                                   |
 | `OBSIDIAN_API_KEY`              | _(unset)_                  | API key for Obsidian Vault backend (used by `genericBackend.ts` known backend preset).                        |

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

@@ -30,17 +30,25 @@ export const opencodeProvider: RegistryEntry = {
     // content (see issue #10867). The opencode provider is passthrough, so
     // declaring them here only sets the wire format / capability flags — the
     // live upstream model list already advertises both ids.
+    // #12681: real window confirmed against the opencode-go registry's own
+    // muse-spark-1.2-contributor entries (contextLength: 1048576, maxOutputTokens:
+    // 131072) — without an explicit value here resolution fell back to the
+    // provider-wide defaultContextLength (200000), understating the real window.
     {
       id: "muse-spark-1.2",
       name: "Muse Spark 1.2",
       supportsReasoning: true,
       targetFormat: "openai-responses",
+      contextLength: 1048576,
+      maxOutputTokens: 131072,
     },
     {
       id: "muse-spark-1.2-contributor-free",
       name: "Muse Spark 1.2 Contributor Free",
       supportsReasoning: true,
       targetFormat: "openai-responses",
+      contextLength: 1048576,
+      maxOutputTokens: 131072,
     },
     { id: "deepseek-v4-flash-free", name: "DeepSeek V4 Flash Free", supportsReasoning: true },
     // #6998: 2026-07-14 refresh — the upstream free tier rotated its lineup;

open-sse/config/providers/registry/opencode/zen/index.ts

@@ -63,11 +63,17 @@ export const opencode_zenProvider: RegistryEntry = {
     // targetFormat declaration, so requests routed here still hit
     // /chat/completions with a mismatched or unanswerable body and the
     // upstream returns an empty message.
+    // #12681: real window confirmed against the opencode-go registry's own
+    // muse-spark-1.2-contributor entries (contextLength: 1048576, maxOutputTokens:
+    // 131072) — without an explicit value here resolution fell back to the
+    // provider-wide defaultContextLength (200000), understating the real window.
     {
       id: "muse-spark-1.2",
       name: "Muse Spark 1.2",
       supportsReasoning: true,
       targetFormat: "openai-responses",
+      contextLength: 1048576,
+      maxOutputTokens: 131072,
     },
     // Explicit wire-format overlay of the base opencode provider's muse-spark entry
     // (targetFormat: openai-responses). Keep in sync with base on catalog syncs.
@@ -76,6 +82,8 @@ export const opencode_zenProvider: RegistryEntry = {
       name: "Muse Spark 1.2 Contributor Free",
       supportsReasoning: true,
       targetFormat: "openai-responses",
+      contextLength: 1048576,
+      maxOutputTokens: 131072,
     },
 
     // ── DeepSeek ────────────────────────────────────────────────

open-sse/executors/opencode.ts

@@ -31,6 +31,13 @@ import {
 import { isOpencodeGeoBlocked, proxyKeyOf } from "./opencodeGeoBlock.ts";
 import { isNetworkRotationSharedEgressGuardEnabled } from "@/shared/utils/featureFlags";
 
+/**
+ * The main OpenCode Zen host, shared by the `opencode` and `opencode-zen`
+ * registry entries. Used to scope the `x-api-key` auth override (#12633) away
+ * from `opencode-go`, which serves a different upstream (`.../zen/go/v1`).
+ */
+const ZEN_BASE_URL = "https://opencode.ai/zen/v1";
+
 /**
  * Per-account proxy configuration, persisted by NoAuthAccountCard under
  * `providerSpecificData.accountProxies` (keyed by the account id, which the UI
@@ -776,6 +783,20 @@ export class OpencodeExecutor extends BaseExecutor {
     }
   }
 
+  /**
+   * #12633: OpenCode Zen's `/v1/responses` endpoint (reached when
+   * `_requestFormat === "openai-responses"`, e.g. Muse Spark Contributor
+   * models) requires `x-api-key`, not `Authorization: Bearer` — unlike the
+   * default `/chat/completions` endpoint on the same host, which accepts
+   * Bearer. Scoped by baseUrl (not provider id/alias) so this only applies to
+   * the main Zen host (`opencode` / `opencode-zen`, both `https://opencode.ai/zen/v1`)
+   * and never to opencode-go, which serves Responses-format models from a
+   * different upstream (`https://opencode.ai/zen/go/v1`) that expects Bearer.
+   */
+  private usesZenApiKeyAuth(): boolean {
+    return this._requestFormat === "openai-responses" && this.config?.baseUrl === ZEN_BASE_URL;
+  }
+
   buildHeaders(
     credentials: ProviderCredentials | null,
     stream = true,
@@ -792,7 +813,7 @@ export class OpencodeExecutor extends BaseExecutor {
       : undefined;
 
     if (key) {
-      if (this._requestFormat === "claude") {
+      if (this._requestFormat === "claude" || this.usesZenApiKeyAuth()) {
         headers["x-api-key"] = key;
       } else {
         headers["Authorization"] = `Bearer ${key}`;

src/lib/memory/__tests__/rerank-loopback-auth-12745.test.ts

@@ -1,214 +0,0 @@
-/**
- * src/lib/memory/__tests__/rerank-loopback-auth-12745.test.ts
- *
- * Regression guard for #12745 — applyRerank()'s internal loopback call to
- * /v1/rerank used to carry no credential, so with REQUIRE_API_KEY=true the
- * global authz proxy's clientApiPolicy would 401 it and rerank silently
- * degraded to unranked order (fail-open by design, so nothing ever surfaced
- * the failure).
- *
- * This file proves two things:
- *   1. The loopback fetch retrieval.ts's applyRerank() issues now carries a
- *      real Authorization: Bearer <internal key> header (fixed by attaching
- *      pickApiKeyForInternalUse() — the same internal-probe selector already
- *      used by combo-health-check / cloud-sync-verify).
- *   2. That fix was NOT done by exempting /v1/rerank from auth: an
- *      unauthenticated *external* request to /api/v1/rerank is still
- *      rejected by clientApiPolicy when REQUIRE_API_KEY=true.
- */
-
-import { describe, test, expect, vi, beforeEach, afterEach } from "vitest";
-import fs from "node:fs";
-import os from "node:os";
-import path from "node:path";
-
-const TEST_DATA_DIR = fs.mkdtempSync(path.join(os.tmpdir(), "omr-rerank-auth-12745-"));
-process.env.DATA_DIR = TEST_DATA_DIR;
-process.env.DISABLE_SQLITE_AUTO_BACKUP = "true";
-process.env.VECTOR_STORE_DISABLE_VEC = "true";
-
-const INTERNAL_KEY = "sk-internal-test-key-12745";
-
-vi.mock("../settings", () => ({
-  getMemorySettings: async () => ({
-    enabled: true,
-    maxTokens: 2000,
-    retentionDays: 30,
-    strategy: "semantic",
-    skillsEnabled: true,
-    embeddingSource: "static",
-    embeddingProviderModel: null,
-    customBaseUrl: null,
-    customModelId: null,
-    transformersEnabled: false,
-    staticEnabled: true,
-    rerankEnabled: true,
-    rerankProviderModel: "test-provider/test-rerank-model",
-    vectorStore: "sqlite-vec",
-    primaryBackend: "sqlite",
-    fallbackBackends: [],
-    backendConfigs: {},
-  }),
-}));
-
-vi.mock("../embedding", () => ({
-  resolveEmbeddingSource: () => ({
-    source: "static",
-    model: "static-hash-8",
-    dimensions: 8,
-    identity: "static",
-    signature: "static-8",
-    reason: "test: static embedding, no network",
-  }),
-  embed: async () => ({
-    vector: new Float32Array([1, 0, 0, 0, 0, 0, 0, 0]),
-    source: "static",
-    model: "static-hash-8",
-    dimensions: 8,
-    latencyMs: 0,
-  }),
-}));
-
-vi.mock("../vectorStore", () => ({
-  getVectorStore: () => ({
-    ensureReady: async () => ({ ready: true, reason: "test" }),
-    upsertVector: async () => undefined,
-    deleteVector: async () => undefined,
-    searchVector: async () => [
-      { memoryId: "rrk-auth-1", score: 0.91 },
-      { memoryId: "rrk-auth-2", score: 0.82 },
-    ],
-    searchHybrid: async () => [],
-    stats: async () => ({ rowCount: 2, needsReindex: 0, activeDim: 8 }),
-  }),
-}));
-
-vi.mock("../../db/apiKeys", async (importOriginal) => {
-  const actual = await importOriginal<typeof import("../../db/apiKeys")>();
-  return {
-    ...actual,
-    pickApiKeyForInternalUse: vi.fn(async () => INTERNAL_KEY),
-  };
-});
-
-const core = await import("../../db/core");
-const { retrievePreview } = await import("../retrieval");
-const { pickApiKeyForInternalUse } = await import("../../db/apiKeys");
-
-function cleanupDb() {
-  core.resetDbInstance();
-  fs.rmSync(TEST_DATA_DIR, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 });
-  fs.mkdirSync(TEST_DATA_DIR, { recursive: true });
-}
-
-function insertMemory(apiKeyId: string, id: string, content: string) {
-  const db = core.getDbInstance();
-  db.prepare(
-    `INSERT INTO memories (id, api_key_id, session_id, type, key, content, metadata, created_at, updated_at, expires_at)
-     VALUES (?, ?, ?, 'factual', ?, ?, '{}', datetime('now'), datetime('now'), NULL)`
-  ).run(id, apiKeyId, "", `key-${id}`, content);
-}
-
-let originalFetch: typeof globalThis.fetch;
-
-beforeEach(() => {
-  cleanupDb();
-  originalFetch = globalThis.fetch;
-  vi.mocked(pickApiKeyForInternalUse).mockClear();
-});
-
-afterEach(() => {
-  globalThis.fetch = originalFetch;
-});
-
-describe("#12745 — memory rerank loopback call authentication", () => {
-  test("applyRerank()'s loopback fetch to /v1/rerank carries an internal Authorization bearer", async () => {
-    insertMemory("api-rrk-auth", "rrk-auth-1", "The capital of France is Paris.");
-    insertMemory("api-rrk-auth", "rrk-auth-2", "TypeScript is a superset of JavaScript.");
-
-    const calls: Array<{ url: string; headers: Record<string, string> }> = [];
-
-    globalThis.fetch = vi.fn(async (input: RequestInfo | URL, init?: RequestInit) => {
-      const url = typeof input === "string" ? input : input.toString();
-      const headers: Record<string, string> = {};
-      new Headers(init?.headers).forEach((value, key) => {
-        headers[key.toLowerCase()] = value;
-      });
-      calls.push({ url, headers });
-
-      // Emulate the REAL clientApiPolicy behavior this loopback call hits in
-      // production: reject without a bearer/x-api-key, accept a valid one.
-      const hasCredential = Boolean(headers["authorization"] || headers["x-api-key"]);
-      if (!hasCredential) {
-        return new Response(JSON.stringify({ error: { message: "Authentication required" } }), {
-          status: 401,
-        });
-      }
-      return new Response(
-        JSON.stringify({
-          results: [
-            { index: 1, relevance_score: 0.95 },
-            { index: 0, relevance_score: 0.4 },
-          ],
-        }),
-        { status: 200 }
-      );
-    }) as unknown as typeof globalThis.fetch;
-
-    const bundle = await retrievePreview("api-rrk-auth", "capital of France", {
-      strategy: "semantic",
-      maxTokens: 2000,
-      limit: 5,
-    });
-
-    expect(calls.length).toBeGreaterThan(0);
-    const rerankCall = calls.find((c) => c.url.includes("/v1/rerank"));
-    expect(rerankCall).toBeDefined();
-
-    const hasCredential = Boolean(
-      rerankCall?.headers["authorization"] || rerankCall?.headers["x-api-key"]
-    );
-    expect(hasCredential).toBe(true);
-    expect(rerankCall?.headers["authorization"]).toBe(`Bearer ${INTERNAL_KEY}`);
-
-    // Functional consequence: with a valid credential the rerank response is
-    // actually honored (item order follows relevance_score) instead of
-    // silently keeping pre-rerank vector-search order.
-    expect(bundle.items[0]?.memory.id).toBe("rrk-auth-2");
-  });
-
-  test("without a credential the same loopback call would still be 401'd (no auth bypass introduced)", async () => {
-    insertMemory("api-rrk-noauth", "rrk-auth-1", "The capital of France is Paris.");
-    insertMemory("api-rrk-noauth", "rrk-auth-2", "TypeScript is a superset of JavaScript.");
-
-    // Simulate the pre-fix condition: internal key selector finds nothing.
-    vi.mocked(pickApiKeyForInternalUse).mockResolvedValueOnce(null);
-
-    let sawUnauthenticatedRerankCall = false;
-    globalThis.fetch = vi.fn(async (input: RequestInfo | URL, init?: RequestInit) => {
-      const url = typeof input === "string" ? input : input.toString();
-      const headers: Record<string, string> = {};
-      new Headers(init?.headers).forEach((value, key) => {
-        headers[key.toLowerCase()] = value;
-      });
-      const hasCredential = Boolean(headers["authorization"] || headers["x-api-key"]);
-      if (url.includes("/v1/rerank") && !hasCredential) {
-        sawUnauthenticatedRerankCall = true;
-        return new Response(JSON.stringify({ error: { message: "Authentication required" } }), {
-          status: 401,
-        });
-      }
-      return new Response(JSON.stringify({ results: [] }), { status: 200 });
-    }) as unknown as typeof globalThis.fetch;
-
-    const bundle = await retrievePreview("api-rrk-noauth", "capital of France", {
-      strategy: "semantic",
-      maxTokens: 2000,
-      limit: 5,
-    });
-
-    expect(sawUnauthenticatedRerankCall).toBe(true);
-    // Fail-open by design: retrieval keeps working (unranked) rather than throwing.
-    expect(bundle.items.length).toBe(2);
-  });
-});

src/lib/memory/retrieval.ts

@@ -12,8 +12,6 @@ import { getQdrantConfig, checkQdrantHealth, searchSemanticMemory } from "./qdra
 import type { MemoryEngineStatus } from "@/shared/schemas/memory";
 import { supportsFts5 } from "../db/migrationRunner";
 import type { SqliteAdapter } from "../db/adapters/types";
-import { pickApiKeyForInternalUse } from "../db/apiKeys";
-import { getRuntimePorts } from "../runtime/ports";
 import {
   estimateTokens,
   parseMetadata,
@@ -145,29 +143,16 @@ function buildFtsRows(apiKeyId: string, config: FtsColConfig): MemoryRow[] {
   }
 }
 
-// Loopback rerank URL — localhost only, never routed over the network. The port is
-// derived from the same runtime source every other internal self-call uses
-// (getRuntimePorts()/process.env.PORT — see src/lib/runtime/ports.ts), never hardcoded,
-// so this keeps working when an operator overrides PORT/API_PORT (#12745).
-function getRerankLoopbackUrl(): string {
-  const { apiPort } = getRuntimePorts();
-  // nosemgrep: javascript.lang.security.audit.non-literal-regexp.non-literal-regexp
-  return `http://127.0.0.1:${apiPort}/v1/rerank`;
-}
+// Loopback rerank URL — localhost only, never routed over the network.
+// nosemgrep: javascript.lang.security.audit.non-literal-regexp.non-literal-regexp
+const RERANK_LOOPBACK_URL = "http://127.0.0.1:20128/v1/rerank";
 
 /**
  * Apply reranking via /v1/rerank (loopback-only) if rerankEnabled + rerankProviderModel is set.
  * Returns reordered array (or original order on any error — rerank failure never fails retrieval).
  *
- * Auth note (#12745): /v1/rerank is a CLIENT_API route gated by clientApiPolicy — with
- * REQUIRE_API_KEY=true an unauthenticated loopback call gets 401'd by the same policy
- * that protects it from the outside, and this call used to send no credential at all,
- * silently degrading retrieval to unranked order. Attach a real, DB-backed API key
- * (the same internal-probe selector already used by combo-health-check / cloud-sync-verify,
- * see pickApiKeyForInternalUse()) as a Bearer token instead of exempting the route.
- *
- * Security note: the URL is a loopback address (127.0.0.1) — it never carries sensitive
- * data over a network link. HTTP is safe for loopback-only IPC.
+ * Security note: the URL is a hardcoded loopback address (127.0.0.1:20128) — it never
+ * carries sensitive data over a network link. HTTP is safe for loopback-only IPC.
  * nosemgrep: javascript.lang.security.detect-non-literal-url
  */
 async function applyRerank<T extends { memory: Memory; score: number }>(
@@ -186,14 +171,10 @@ async function applyRerank<T extends { memory: Memory; score: number }>(
       top_n: items.length,
     };
 
-    const internalKey = await pickApiKeyForInternalUse("internal-probe");
-    const headers: Record<string, string> = { "content-type": "application/json" };
-    if (internalKey) headers.authorization = `Bearer ${internalKey}`;
-
-    const res = await fetch(getRerankLoopbackUrl(), {
+    const res = await fetch(RERANK_LOOPBACK_URL, {
       // nosemgrep: typescript.react.security.react-insecure-request.react-insecure-request
       method: "POST",
-      headers,
+      headers: { "content-type": "application/json" },
       body: JSON.stringify(body),
       signal: AbortSignal.timeout(5000),
     });

tests/unit/issue-12633-opencode-zen-responses-auth.test.ts

@@ -0,0 +1,54 @@
+import test from "node:test";
+import assert from "node:assert/strict";
+import { OpencodeExecutor } from "../../open-sse/executors/opencode.ts";
+
+test("#12633: openai-responses format on opencode-zen sends x-api-key, not Authorization Bearer", () => {
+  const executor = new OpencodeExecutor("opencode-zen");
+  executor._requestFormat = "openai-responses";
+  const headers = executor.buildHeaders(
+    { apiKey: "sk-zen-test" },
+    true,
+    null,
+    "muse-spark-1.2-contributor-free"
+  );
+
+  assert.equal(headers["x-api-key"], "sk-zen-test");
+  assert.equal(headers["Authorization"], undefined);
+});
+
+test("#12633: openai-responses format on the base opencode (oc) provider also sends x-api-key", () => {
+  const executor = new OpencodeExecutor("opencode");
+  executor._requestFormat = "openai-responses";
+  const headers = executor.buildHeaders(
+    { apiKey: "sk-oc-test" },
+    true,
+    null,
+    "muse-spark-1.2-contributor-free"
+  );
+
+  assert.equal(headers["x-api-key"], "sk-oc-test");
+  assert.equal(headers["Authorization"], undefined);
+});
+
+test("#12633: openai-responses format on opencode-go (different upstream endpoint) keeps Authorization Bearer", () => {
+  const executor = new OpencodeExecutor("opencode-go");
+  executor._requestFormat = "openai-responses";
+  const headers = executor.buildHeaders(
+    { apiKey: "sk-go-test" },
+    true,
+    null,
+    "muse-spark-1.2-contributor"
+  );
+
+  assert.equal(headers["Authorization"], "Bearer sk-go-test");
+  assert.equal(headers["x-api-key"], undefined);
+});
+
+test("#12633: claude format keeps sending x-api-key (unchanged behavior)", () => {
+  const executor = new OpencodeExecutor("opencode-zen");
+  executor._requestFormat = "claude";
+  const headers = executor.buildHeaders({ apiKey: "sk-claude-test" }, true, null, "some-model");
+
+  assert.equal(headers["x-api-key"], "sk-claude-test");
+  assert.equal(headers["Authorization"], undefined);
+});

tests/unit/issue-12681-opencode-muse-spark-context.test.ts

@@ -0,0 +1,33 @@
+import { test } from "node:test";
+import assert from "node:assert/strict";
+import { REGISTRY } from "../../open-sse/config/providerRegistry.ts";
+import { getTokenLimit } from "../../open-sse/services/contextManager.ts";
+
+test("#12681: opencode registry declares an explicit real contextLength for muse-spark-1.2 models", () => {
+  const opencode = REGISTRY["opencode"];
+  const museSpark = opencode.models.find((m) => m.id === "muse-spark-1.2");
+  const museSparkFree = opencode.models.find((m) => m.id === "muse-spark-1.2-contributor-free");
+  assert.notEqual(
+    museSpark?.contextLength,
+    undefined,
+    "muse-spark-1.2 should declare its own real contextLength instead of relying on the 200000 provider default"
+  );
+  assert.notEqual(
+    museSparkFree?.contextLength,
+    undefined,
+    "muse-spark-1.2-contributor-free should declare its own real contextLength instead of relying on the 200000 provider default"
+  );
+});
+
+test("#12681: opencode-zen registry declares an explicit real contextLength for muse-spark-1.2 models", () => {
+  const zen = REGISTRY["opencode-zen"];
+  const museSpark = zen.models.find((m) => m.id === "muse-spark-1.2");
+  const museSparkFree = zen.models.find((m) => m.id === "muse-spark-1.2-contributor-free");
+  assert.notEqual(museSpark?.contextLength, undefined);
+  assert.notEqual(museSparkFree?.contextLength, undefined);
+});
+
+test("#12681: contextManager.getTokenLimit resolves muse-spark-1.2-contributor-free to its real 1M+ window, not the 200000 provider default", () => {
+  assert.equal(getTokenLimit("opencode", "muse-spark-1.2-contributor-free"), 1048576);
+  assert.equal(getTokenLimit("opencode-zen", "muse-spark-1.2-contributor-free"), 1048576);
+});