fix: auto-mode stops after provider errors (#2762) (#2764)

* feat: Registered 6 MCP tools (gsd_execute, gsd_status, gsd_result, gsd_…

- "packages/mcp-server/src/server.ts"
- "packages/mcp-server/src/cli.ts"
- "packages/mcp-server/src/index.ts"
- "packages/rpc-client/dist/index.d.ts"

GSD-Task: S05/T02

* docs: Added 31 integration tests, build pipeline, and consumer README f…

- "packages/mcp-server/src/mcp-server.test.ts"
- "packages/mcp-server/README.md"
- "packages/mcp-server/dist/"

GSD-Task: S05/T03

* fix: prevent auto-mode hard stop on provider errors and suppress duplicate async_job_result follow-ups (#2762)

Two compounding bugs caused auto-mode to silently die after unit completion:

1. async_job_result follow-ups fired after unit completion because deliverResult
   ran synchronously in the job promise .then() chain, racing with await_job's
   .then() that sets job.awaited=true. Deferring delivery by one microtask via
   queueMicrotask ensures await_job marks the job first.

2. Provider error pause converted to hard stop because pauseAuto resolved the
   unit promise with {status:"cancelled"} but no ErrorContext, so runUnitPhase
   treated it identically to a session-creation timeout and called stopAuto.
   Now pauseAuto accepts and forwards ErrorContext, and runUnitPhase checks for
   category:"provider" to break without hard-stopping.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* test: update source-scan assertion for new pauseAuto signature

The structural test checked for `resolveAgentEndCancelled()` with empty
parens. Now that pauseAuto passes _errorContext, match the call prefix.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

src/resources/extensions/async-jobs/job-manager.ts

@@ -172,7 +172,10 @@ export class AsyncJobManager {
 
 	private deliverResult(job: Job): void {
 		if (!this.onJobComplete) return;
-		this.onJobComplete(job);
+		// Defer delivery by one microtask so await_job's .then() chain runs first
+		// and can set job.awaited = true before onJobComplete checks it (#2762).
+		const cb = this.onJobComplete;
+		queueMicrotask(() => cb(job));
 	}
 
 	private scheduleEviction(id: string): void {

src/resources/extensions/gsd/auto.ts

@@ -186,7 +186,7 @@ import {
   postUnitPostVerification,
 } from "./auto-post-unit.js";
 import { bootstrapAutoSession, type BootstrapDeps } from "./auto-start.js";
-import { autoLoop, resolveAgentEnd, resolveAgentEndCancelled, _resetPendingResolve, isSessionSwitchInFlight, type LoopDeps } from "./auto-loop.js";
+import { autoLoop, resolveAgentEnd, resolveAgentEndCancelled, _resetPendingResolve, isSessionSwitchInFlight, type LoopDeps, type ErrorContext } from "./auto-loop.js";
 import {
   WorktreeResolver,
   type WorktreeResolverDeps,
@@ -800,11 +800,14 @@ export async function stopAuto(
 export async function pauseAuto(
   ctx?: ExtensionContext,
   _pi?: ExtensionAPI,
+  _errorContext?: ErrorContext,
 ): Promise<void> {
   if (!s.active) return;
   clearUnitTimeout();
   // Unblock any pending unit promise so the auto-loop is not orphaned.
-  resolveAgentEndCancelled();
+  // Pass errorContext so runUnitPhase can distinguish user-initiated pause
+  // from provider-error pause and avoid hard-stopping (#2762).
+  resolveAgentEndCancelled(_errorContext);
 
   s.pausedSessionFile = ctx?.sessionManager?.getSessionFile() ?? null;
 

src/resources/extensions/gsd/auto/phases.ts

@@ -1069,6 +1069,12 @@ export async function runUnitPhase(
   }
 
   if (unitResult.status === "cancelled") {
+    // Provider-error pause: pauseAuto already handled cleanup and scheduled
+    // recovery. Don't hard-stop — just break out of the loop (#2762).
+    if (unitResult.errorContext?.category === "provider") {
+      debugLog("autoLoop", { phase: "exit", reason: "provider-pause", isTransient: unitResult.errorContext.isTransient });
+      return { action: "break", reason: "provider-pause" };
+    }
     ctx.ui.notify(
       `Session creation timed out or was cancelled for ${unitType} ${unitId}. Will retry.`,
       "warning",

src/resources/extensions/gsd/bootstrap/agent-end-recovery.ts

@@ -33,7 +33,12 @@ async function pauseTransientWithBackoff(
   if (!allowAutoResume) {
     ctx.ui.notify(`Transient provider errors persisted after ${MAX_TRANSIENT_AUTO_RESUMES} auto-resume attempts. Pausing for manual review.`, "warning");
   }
-  await pauseAutoForProviderError(ctx.ui, errorDetail, () => pauseAuto(ctx, pi), {
+  await pauseAutoForProviderError(ctx.ui, errorDetail, () => pauseAuto(ctx, pi, {
+    message: `Provider error: ${errorDetail}`,
+    category: "provider",
+    isTransient: allowAutoResume,
+    retryAfterMs,
+  }), {
     isRateLimit,
     isTransient: allowAutoResume,
     retryAfterMs,
@@ -161,7 +166,11 @@ export async function handleAgentEnd(
     }
 
     // --- Permanent / unknown: pause indefinitely ---
-    await pauseAutoForProviderError(ctx.ui, errorDetail, () => pauseAuto(ctx, pi), {
+    await pauseAutoForProviderError(ctx.ui, errorDetail, () => pauseAuto(ctx, pi, {
+      message: `Provider error: ${errorDetail}`,
+      category: "provider",
+      isTransient: false,
+    }), {
       isRateLimit: false,
       isTransient: false,
       retryAfterMs: 0,

src/resources/extensions/gsd/tests/agent-end-retry.test.ts

@@ -102,7 +102,7 @@ test("pauseAuto calls resolveAgentEndCancelled to unblock the loop", () => {
   const fnBlock = source.slice(fnIdx, source.indexOf("\n/**\n * Build", fnIdx + 100));
 
   assert.ok(
-    fnBlock.includes("resolveAgentEndCancelled()"),
+    fnBlock.includes("resolveAgentEndCancelled("),
     "pauseAuto must call resolveAgentEndCancelled to unblock the auto-loop promise",
   );
 });