From abd865d4f2c6b5d9d89057b4a11fc731a46f1954 Mon Sep 17 00:00:00 2001
From: Aditya kumar singh <143548997+Adityakk9031@users.noreply.github.com>
Date: Thu, 16 Jul 2026 15:41:04 +0530
Subject: [PATCH] fix(run-queue): handle worker spawn failure gracefully and
 avoid raw tracebacks

---
 .../workflows/run_queue/service.py            | 19 +++++++---
 tests/test_run_queue_workflow.py              | 35 +++++++++++++++++++
 2 files changed, 50 insertions(+), 4 deletions(-)

diff --git a/src/codealmanac/workflows/run_queue/service.py b/src/codealmanac/workflows/run_queue/service.py
index 71eeb242..53074885 100644
--- a/src/codealmanac/workflows/run_queue/service.py
+++ b/src/codealmanac/workflows/run_queue/service.py
@@ -1,6 +1,6 @@
 from pathlib import Path
 
-from codealmanac.core.errors import error_summary
+from codealmanac.core.errors import ExecutionFailed, error_summary
 from codealmanac.services.repositories.service import RepositoriesService
 from codealmanac.services.runs.models import (
     RunCancelResult,
@@ -86,7 +86,7 @@ def queue_build(self, request: BuildRequest) -> RunRecord:
 
     def start_build(self, request: BuildRequest) -> RunQueueStartResult:
         run = self.queue_build(request)
-        worker = self.spawn_worker(request.path)
+        worker = self._spawn_worker_safely(run, request.path)
         return self.start_result(run, worker)
 
     def queue_ingest(self, request: IngestRequest) -> RunRecord:
@@ -104,7 +104,7 @@ def queue_ingest(self, request: IngestRequest) -> RunRecord:
 
     def start_ingest(self, request: IngestRequest) -> RunQueueStartResult:
         run = self.queue_ingest(request)
-        worker = self.spawn_worker(request.cwd)
+        worker = self._spawn_worker_safely(run, request.cwd)
         return self.start_result(run, worker)
 
     def queue_garden(self, request: GardenRequest) -> RunRecord:
@@ -122,7 +122,7 @@ def queue_garden(self, request: GardenRequest) -> RunRecord:
 
     def start_garden(self, request: GardenRequest) -> RunQueueStartResult:
         run = self.queue_garden(request)
-        worker = self.spawn_worker(request.cwd)
+        worker = self._spawn_worker_safely(run, request.cwd)
         return self.start_result(run, worker)
 
     def start_scheduled_garden(
@@ -166,6 +166,17 @@ def start_scheduled_garden(
     def spawn_worker(self, cwd: Path) -> RunWorkerSpawnResult:
         return self.spawner.spawn(SpawnRunWorkerRequest(cwd=cwd))
 
+    def _spawn_worker_safely(self, run: RunRecord, cwd: Path) -> RunWorkerSpawnResult:
+        try:
+            return self.spawn_worker(cwd)
+        except Exception as error:
+            raise ExecutionFailed(
+                f"Run queued as {run.run_id} but worker failed to start: {error}. "
+                f"Use 'codealmanac jobs attach {run.run_id}' "
+                "once a worker is available."
+            ) from error
+
+
     def start_result(
         self,
         run: RunRecord,
diff --git a/tests/test_run_queue_workflow.py b/tests/test_run_queue_workflow.py
index 5598281e..91ee4a99 100644
--- a/tests/test_run_queue_workflow.py
+++ b/tests/test_run_queue_workflow.py
@@ -119,6 +119,41 @@ def test_run_queue_start_persists_spec_and_spawns_worker(
     assert not (repo / "almanac/jobs").exists()
 
 
+class FailingWorkerSpawner:
+    def spawn(self, request: SpawnRunWorkerRequest) -> RunWorkerSpawnResult:
+        raise OSError("fake spawn failure")
+
+
+def test_run_queue_start_handles_worker_spawn_failure(
+    tmp_path: Path,
+    isolated_home: Path,
+):
+    repo = tmp_path / "repo"
+    repo.mkdir()
+    (repo / "note.md").write_text("queue design note\n", encoding="utf-8")
+    spawner = FailingWorkerSpawner()
+    app = create_app(
+        AppConfig(database_path=isolated_home / ".codealmanac/codealmanac.db"),
+        harness_adapters=(QueueWritingHarnessAdapter(),),
+        worker_spawner=spawner,
+    )
+    initialize_repository(app, repo)
+
+    with pytest.raises(ExecutionFailed, match="fake spawn failure"):
+        app.workflows.queue.start_ingest(
+            IngestRequest(
+                cwd=repo,
+                inputs=("note.md",),
+                harness=HarnessKind.CODEX,
+                model="gpt-5.5",
+            )
+        )
+
+    runs = app.runs.list(ListRunsRequest(repository_name=repo.name))
+    assert len(runs) == 1
+    assert runs[0].status == RunStatus.QUEUED
+
+
 def test_run_queue_drains_persisted_ingest_spec(
     tmp_path: Path,
     isolated_home: Path,