fix: keep session takeover out of model fallback

cael-dandelion-cult · cael-dandelion-cult · commit f95dc0a85649 · 2026-05-17T21:35:47.000-07:00
diff --git a/src/agents/model-fallback.test.ts b/src/agents/model-fallback.test.ts
@@ -897,6 +897,29 @@ describe("runWithModelFallback", () => {
     expect(errorCall.error).toBe(unknownError);
   });
 
+  it("rethrows embedded session takeover errors instead of walking model fallbacks", async () => {
+    const cfg = makeCfg();
+    const takeoverError = Object.assign(
+      new Error("session file changed while embedded prompt lock was released: /tmp/session.jsonl"),
+      { name: "EmbeddedAttemptSessionTakeoverError" },
+    );
+    const run = vi.fn().mockRejectedValue(takeoverError);
+    const onError = vi.fn();
+
+    await expect(
+      runWithModelFallback({
+        cfg,
+        provider: "openai",
+        model: "gpt-4.1-mini",
+        run,
+        onError,
+      }),
+    ).rejects.toBe(takeoverError);
+
+    expect(run).toHaveBeenCalledTimes(1);
+    expect(onError).not.toHaveBeenCalled();
+  });
+
   it("throws unrecognized error on last candidate", async () => {
     const cfg = makeCfg();
     const run = vi.fn().mockRejectedValueOnce(new Error("something weird"));
diff --git a/src/agents/model-fallback.ts b/src/agents/model-fallback.ts
@@ -48,6 +48,7 @@ import {
 import { isLikelyContextOverflowError } from "./pi-embedded-helpers/errors.js";
 import type { FailoverReason } from "./pi-embedded-helpers/types.js";
 import { resolveSessionSuspensionReason, suspendSession } from "./session-suspension.js";
+import { isSessionWriteLockTimeoutError } from "./session-write-lock-error.js";
 
 const log = createSubsystemLogger("model-fallback");
 
@@ -116,6 +117,23 @@ function shouldRethrowAbort(err: unknown): boolean {
   return isFallbackAbortError(err) && !isTimeoutError(err);
 }
 
+function isEmbeddedAttemptSessionTakeoverError(err: unknown): boolean {
+  if (!err || typeof err !== "object") {
+    return false;
+  }
+  const name = "name" in err ? String(err.name) : "";
+  if (name === "EmbeddedAttemptSessionTakeoverError") {
+    return true;
+  }
+  return formatErrorMessage(err).includes(
+    "session file changed while embedded prompt lock was released",
+  );
+}
+
+function shouldRethrowSessionOwnershipError(err: unknown): boolean {
+  return isEmbeddedAttemptSessionTakeoverError(err) || isSessionWriteLockTimeoutError(err);
+}
+
 function createModelCandidateCollector(allowlist: Set<string> | null | undefined): {
   candidates: ModelCandidate[];
   addExplicitCandidate: (candidate: ModelCandidate) => void;
@@ -231,6 +249,9 @@ async function runFallbackCandidate<T>(params: {
     if (isCommandLaneTaskTimeoutError(err)) {
       throw err;
     }
+    if (shouldRethrowSessionOwnershipError(err)) {
+      throw err;
+    }
     // Normalize abort-wrapped rate-limit errors (e.g. Google Vertex RESOURCE_EXHAUSTED)
     // so they become FailoverErrors and continue the fallback loop instead of aborting.
     const normalizedFailover = coerceToFailoverError(err, {
