fix(opencode): route Muse Spark through the Responses API

muse-spark-1.2-contributor-free returned HTTP 500 on /zen/v1/chat/completions.
The model is only served by /zen/v1/responses, so route it there via a per-model
targetFormat and normalize the Chat fields the Responses API rejects
(max_tokens -> max_output_tokens, reasoning_effort -> reasoning{effort,summary}),
clamping max/ultra down to the highest effort the model accepts (xhigh).

Routing stays per-model: the other free models (big-pickle, hy3-free, mimo,
nemotron, laguna) are not served by /responses and keep /chat/completions.

open-sse/executors/opencode.js

@@ -1,11 +1,13 @@
 import crypto from "crypto";
 import { BaseExecutor } from "./base.js";
 import { PROVIDERS } from "../config/providers.js";
+import { getThinkingLevels } from "../providers/thinkingLevels.js";
 import { injectReasoningContent } from "../utils/reasoningContentInjector.js";
 import { resolveSessionId } from "../utils/sessionManager.js";
 
 const OPENCODE_UA = "opencode";
-const MESSAGES_MODELS = new Set();
+// Models served by /zen/v1/responses; every other model stays on /chat/completions.
+const RESPONSES_MODELS = new Set(["muse-spark-1.2-contributor-free"]);
 
 function generateRequestId() {
   return `msg_${crypto.randomUUID().replace(/-/g, "")}`;
@@ -15,19 +17,47 @@ function generateSessionId() {
   return `ses_${crypto.randomUUID().replace(/-/g, "")}`;
 }
 
-// Normalize any resolved id into opencode's ses_ format (stable per-conversation)
-function toOpencodeSession(id) {
-  const stripped = String(id || "").replace(/^ses_/, "").replace(/-/g, "");
-  return stripped ? `ses_${stripped}` : null;
+// Strip the thinking suffix "model(level)" so registry lookups hit the base id.
+function baseModelId(model) {
+  return String(model || "").replace(/\([^()]+\)\s*$/, "").trim();
+}
+
+function isResponsesModel(model) {
+  return RESPONSES_MODELS.has(baseModelId(model));
 }
 
 function resolveOpencodeSession(body, credentials) {
-  return toOpencodeSession(resolveSessionId({
-    headers: credentials?.rawHeaders,
+  const headers = credentials?.rawHeaders || {};
+  return resolveSessionId({
+    headers,
     body,
     connectionId: credentials?.connectionId,
     scope: "opencode",
-  }));
+    generate: generateSessionId,
+  });
+}
+
+function normalizeOpencodeReasoning(model, body) {
+  const current = body.reasoning;
+  const currentReasoning = current && typeof current === "object" && !Array.isArray(current)
+    ? current
+    : null;
+  const requestedEffort = typeof body.reasoning_effort === "string"
+    ? body.reasoning_effort
+    : currentReasoning?.effort;
+  if (typeof requestedEffort !== "string") return;
+
+  const cleanModel = baseModelId(model || body.model);
+  const supportedLevels = getThinkingLevels("opencode", cleanModel);
+  let effort = requestedEffort.toLowerCase().trim();
+  if ((effort === "max" || effort === "ultra") && supportedLevels?.length && !supportedLevels.includes(effort)) {
+    if (effort === "ultra" && supportedLevels.includes("max")) effort = "max";
+    else if (supportedLevels.includes("xhigh")) effort = "xhigh";
+  }
+
+  body.reasoning = { ...currentReasoning, effort };
+  if (!body.reasoning.summary) body.reasoning.summary = "auto";
+  delete body.reasoning_effort;
 }
 
 export class OpenCodeExecutor extends BaseExecutor {
@@ -38,13 +68,24 @@ export class OpenCodeExecutor extends BaseExecutor {
 
   transformRequest(model, body, stream, credentials) {
     this._currentSessionId = resolveOpencodeSession(body, credentials);
+    if (isResponsesModel(model)) {
+      // Responses API names the output cap max_output_tokens and takes thinking
+      // as reasoning:{effort,summary} — normalize the Chat fields at this boundary.
+      if (body.max_output_tokens === undefined) {
+        if (body.max_completion_tokens !== undefined) body.max_output_tokens = body.max_completion_tokens;
+        else if (body.max_tokens !== undefined) body.max_output_tokens = body.max_tokens;
+      }
+      delete body.max_tokens;
+      delete body.max_completion_tokens;
+      normalizeOpencodeReasoning(model, body);
+    }
     return injectReasoningContent({ provider: this.provider, model, body });
   }
 
   buildUrl(model) {
     const base = this.config.baseUrl;
-    return MESSAGES_MODELS.has(model)
-      ? `${base}/zen/v1/messages`
+    return isResponsesModel(model)
+      ? `${base}/zen/v1/responses`
       : `${base}/zen/v1/chat/completions`;
   }
 

open-sse/providers/capabilities.js

@@ -125,6 +125,8 @@ export const MODEL_CAPABILITIES = {
   "kimi-for-coding-highspeed": { vision: true, videoInput: true, reasoning: true, thinkingFormat: "kimi", thinkingCanDisable: false, contextWindow: 262144, maxOutput: 65536 },
   "kimi-k2.7-code":    { vision: true, videoInput: true, reasoning: true, thinkingFormat: "kimi", thinkingCanDisable: false, contextWindow: 262144, maxOutput: 65536 },
   "kimi-k2.7-code-highspeed": { vision: true, videoInput: true, reasoning: true, thinkingFormat: "kimi", thinkingCanDisable: false, contextWindow: 262144, maxOutput: 65536 },
+  // OpenCode Free Muse Spark — OpenAI Responses reasoning supports up to xhigh.
+  "muse-spark-1.2-contributor-free": { reasoning: true, thinkingFormat: "openai", contextWindow: 1048576, maxOutput: 131072 },
 };
 
 const KIRO_GPT_5_6_CAPABILITIES = { vision: true, reasoning: true, search: true, thinkingFormat: "openai", contextWindow: 272000, maxOutput: 128000 };

open-sse/providers/registry/opencode.js

@@ -19,7 +19,11 @@ export default {
     },
     noAuth: true,
   },
-  models: [],
+  models: [
+    // Only this model is served by /zen/v1/responses; the rest stay on
+    // /chat/completions, so the format is declared per-model, not per-provider.
+    { id: "muse-spark-1.2-contributor-free", name: "Muse Spark 1.2 Contributor Free", targetFormat: "openai-responses" },
+  ],
   modelsFetcher: { url: "https://opencode.ai/zen/v1/models", type: "opencode-free" },
   passthroughModels: true,
 };

open-sse/translator/request/openai-responses.js

@@ -300,7 +300,16 @@ function buildReasoningInputItem(msg) {
  */
 export function openaiToOpenAIResponsesRequest(model, body, stream, credentials) {
   // Body already in Responses API format (e.g. Cursor CLI calling /chat/completions with input[])
-  if (body.input) return { ...body, model, stream: true };
+  if (body.input) {
+    const out = { ...body, model, stream: true };
+    if (out.max_output_tokens === undefined) {
+      if (out.max_completion_tokens !== undefined) out.max_output_tokens = out.max_completion_tokens;
+      else if (out.max_tokens !== undefined) out.max_output_tokens = out.max_tokens;
+    }
+    delete out.max_tokens;
+    delete out.max_completion_tokens;
+    return out;
+  }
 
   const result = {
     model,
@@ -416,7 +425,13 @@ export function openaiToOpenAIResponsesRequest(model, body, stream, credentials)
 
   // Pass through other relevant fields
   if (body.temperature !== undefined) result.temperature = body.temperature;
-  if (body.max_tokens !== undefined) result.max_tokens = body.max_tokens;
+  if (body.max_output_tokens !== undefined) {
+    result.max_output_tokens = body.max_output_tokens;
+  } else if (body.max_completion_tokens !== undefined) {
+    result.max_output_tokens = body.max_completion_tokens;
+  } else if (body.max_tokens !== undefined) {
+    result.max_output_tokens = body.max_tokens;
+  }
   if (body.top_p !== undefined) result.top_p = body.top_p;
   if (body.reasoning !== undefined) result.reasoning = body.reasoning;
   if (body.reasoning_effort !== undefined) result.reasoning = { effort: body.reasoning_effort, summary: "auto" };

tests/unit/executor-const-guard.test.js

@@ -9,6 +9,7 @@ import { DEFAULT_MAX_TOKENS, DEFAULT_MIN_TOKENS } from "../../open-sse/config/ru
 import mimoFree from "../../open-sse/providers/registry/mimo-free.js";
 import opencode from "../../open-sse/providers/registry/opencode.js";
 import antigravity from "../../open-sse/providers/registry/antigravity.js";
+import { OpenCodeExecutor } from "../../open-sse/executors/opencode.js";
 
 describe("compat base URLs / version", () => {
   it("OPENAI_COMPAT_BASE", () => {
@@ -46,3 +47,36 @@ describe("antigravity retry (intentional change: 429=6, 503=3)", () => {
     expect(antigravity.transport.retry["503"].attempts).toBe(3);
   });
 });
+
+describe("OpenCode Free endpoint routing", () => {
+  const MUSE = "muse-spark-1.2-contributor-free";
+
+  it("declares the Responses format only on the Muse Spark model", () => {
+    expect(opencode.transport.format).toBeUndefined();
+    const muse = opencode.models.find((m) => m.id === MUSE);
+    expect(muse?.targetFormat).toBe("openai-responses");
+  });
+
+  it("routes Muse Spark to /responses and every other model to /chat/completions", () => {
+    const executor = new OpenCodeExecutor();
+    expect(executor.buildUrl(MUSE)).toBe("https://opencode.ai/zen/v1/responses");
+    expect(executor.buildUrl(`${MUSE}(xhigh)`)).toBe("https://opencode.ai/zen/v1/responses");
+    expect(executor.buildUrl("big-pickle")).toBe("https://opencode.ai/zen/v1/chat/completions");
+    expect(executor.buildUrl("hy3-free")).toBe("https://opencode.ai/zen/v1/chat/completions");
+  });
+
+  it("normalizes Chat token/thinking fields only for the Responses model", () => {
+    const executor = new OpenCodeExecutor();
+    const muse = { max_tokens: 4096, reasoning_effort: "high" };
+    executor.transformRequest(MUSE, muse, true, {});
+    expect(muse.max_output_tokens).toBe(4096);
+    expect(muse.max_tokens).toBeUndefined();
+    expect(muse.reasoning).toEqual({ effort: "high", summary: "auto" });
+
+    const chat = { max_tokens: 4096, reasoning_effort: "high" };
+    executor.transformRequest("big-pickle", chat, true, {});
+    expect(chat.max_tokens).toBe(4096);
+    expect(chat.max_output_tokens).toBeUndefined();
+    expect(chat.reasoning_effort).toBe("high");
+  });
+});

tests/unit/opencode-muse-spark-thinking.test.js

@@ -0,0 +1,94 @@
+import { describe, expect, it } from "vitest";
+import { getCapabilitiesForModel } from "../../open-sse/providers/capabilities.js";
+import { PROVIDER_MODELS } from "../../open-sse/config/providerModels.js";
+import { getThinkingLevels } from "../../open-sse/providers/thinkingLevels.js";
+import { FORMATS } from "../../open-sse/translator/formats.js";
+import { OpenCodeExecutor } from "../../open-sse/executors/opencode.js";
+import "../translator/registerAll.js";
+import { translateRequest } from "../../open-sse/translator/index.js";
+
+const MODEL = "muse-spark-1.2-contributor-free";
+const PROVIDER = "opencode";
+
+const input = [{
+  type: "message",
+  role: "user",
+  content: [{ type: "input_text", text: "Think, then answer: 2 + 2?" }],
+}];
+
+describe("OpenCode Free Muse Spark thinking", () => {
+  it("advertises reasoning and the requested model limits", () => {
+    expect(PROVIDER_MODELS.oc?.some((model) => model.id === MODEL)).toBe(true);
+    expect(getCapabilitiesForModel(PROVIDER, MODEL)).toMatchObject({
+      reasoning: true,
+      thinkingFormat: "openai",
+      contextWindow: 1048576,
+      maxOutput: 131072,
+    });
+    expect(getCapabilitiesForModel(PROVIDER, `oc/${MODEL}`)).toMatchObject({
+      reasoning: true,
+      contextWindow: 1048576,
+      maxOutput: 131072,
+    });
+    expect(getThinkingLevels(PROVIDER, MODEL)).toEqual([
+      "none",
+      "minimal",
+      "low",
+      "medium",
+      "high",
+      "xhigh",
+    ]);
+  });
+
+  it("clamps max to xhigh and emits the Responses reasoning shape", () => {
+    const body = {
+      input,
+      reasoning: { effort: "max" },
+      max_tokens: 131072,
+    };
+
+    const out = new OpenCodeExecutor().transformRequest(MODEL, body, true, {
+      connectionId: "opencode-muse-spark-test",
+    });
+
+    expect(out.reasoning).toEqual({ effort: "xhigh", summary: "auto" });
+    expect(out.reasoning_effort).toBeUndefined();
+    expect(out.max_output_tokens).toBe(131072);
+    expect(out.max_tokens).toBeUndefined();
+  });
+
+  it("leaves the other free models on Chat Completions", () => {
+    const executor = new OpenCodeExecutor();
+    const body = { messages: [{ role: "user", content: "hi" }], max_tokens: 1024 };
+    executor.transformRequest("big-pickle", body, true, {});
+    expect(executor.buildUrl("big-pickle")).toBe("https://opencode.ai/zen/v1/chat/completions");
+    expect(body.max_tokens).toBe(1024);
+    expect(body.max_output_tokens).toBeUndefined();
+  });
+
+  it("translates Chat Completions max thinking into a Responses request", () => {
+    const body = {
+      model: `oc/${MODEL}`,
+      messages: [{ role: "user", content: "Think, then answer: 2 + 2?" }],
+      reasoning_effort: "max",
+      max_tokens: 131072,
+    };
+
+    const translated = translateRequest(
+      FORMATS.OPENAI,
+      FORMATS.OPENAI_RESPONSES,
+      MODEL,
+      body,
+      true,
+      {},
+      PROVIDER,
+    );
+    const out = new OpenCodeExecutor().transformRequest(MODEL, translated, true, {
+      connectionId: "opencode-muse-spark-translation-test",
+    });
+
+    expect(out.reasoning).toEqual({ effort: "xhigh", summary: "auto" });
+    expect(out.max_output_tokens).toBe(131072);
+    expect(out.max_tokens).toBeUndefined();
+  });
+});