Files
PaperClipAI/server/src/services/execution-control-reconciliation.ts
T
Devin FoleyandClaude Opus 5.5 09f9d691cf fix(server): keep the root failure code when the execution deadline fails a run (#19028)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work.
> - The execution-control watchdog fails a run when its control deadline
passes and no owner finished it.
> - The watchdog replaces the run's error code with
`execution_finalization_deadline_exceeded`.
> - The run-failure report reads the root cause from
`resultJson.originalFailureCode`. The watchdog did not write that field.
> - So every watchdog failure goes to Sentry with the same fingerprint,
whatever its first failure was.
> - This pull request keeps the earlier failure code as
`originalFailureCode`, as the native finalizer already does.
> - The benefit is that Sentry groups watchdog failures by their real
first failure, so each cause can be found and fixed.

## Linked Issues or Issue Description

**What happened?**

One hosted Sentry group collects all
`execution_finalization_deadline_exceeded` run failures. It has 17
events from many drivers (Codex app server, OpenCode server, ACPX
runtime). In one recent event, the run first failed with
`native_execution_ownership_unverified`. The watchdog then failed it at
the deadline. The recovery action evidence kept `originalFailureCode:
native_execution_ownership_unverified`, but the run row and the Sentry
event did not. The event had no `root_error_code` tag, and its
fingerprint was only `[execution_finalization_deadline_exceeded,
paperclip_runner]`.

**Expected behavior**

When the watchdog replaces an earlier, more specific error code, the run
keeps that code as `resultJson.originalFailureCode`. The run-failure
report then sends `root_error_code` and adds it to the fingerprint
(`server/src/sentry.ts`).

**Steps to reproduce**

1. Create a running run with `errorCode:
native_execution_ownership_unverified` and an expired
`executionControlDeadlineAt`.
2. Run `reconcileAbandonedExecutionControl`.
3. The run fails with `execution_finalization_deadline_exceeded`, and
the earlier code is lost. The captured event has no
`diagnostics.execution.originalFailureCode`.

**Paperclip version or commit**

`master` at `89703c3ff7de`.

Refs #14110, a similar fix for the finalization sweep. A search found no
other open PR for this path. Draft #18672 changes legacy deferred wakes,
which is a different path.

## What Changed

- `execution-control-reconciliation.ts`: when the watchdog fails a run,
it writes the earlier code (`coordinator.failureCode ?? run.errorCode`)
to `resultJson.originalFailureCode`. This is the same value that it
already writes to the recovery action evidence.
- The watchdog keeps the other `resultJson` fields.
- The watchdog does not replace a root cause that is already recorded.
- The watchdog does not record the deadline code as its own root cause.
- Two regression tests in `execution-control-reconciliation.test.ts`.

## Verification

- The new test "keeps the run's prior failure code as the reported root
cause" fails on unchanged `master`. The run's `resultJson` has no
`originalFailureCode`. With the fix, the test passes.
- `pnpm exec vitest run
src/services/execution-control-reconciliation.test.ts
src/__tests__/execution-control-work.test.ts
src/__tests__/run-failure-sentry.test.ts src/__tests__/sentry.test.ts`:
111 passed, 5 skipped.
- `pnpm exec vitest run src/services/task-retry.test.ts
src/services/heartbeat/recovery.test.ts
src/services/native-runtime/native-safe-replacement.test.ts
src/services/workspace-recovery.test.ts`: 437 passed. These suites use
this watchdog cause.
- `node scripts/typecheck-layers.mjs` in `server/`: passed.
- Not run locally: the full `pnpm -r typecheck` (the sandbox has no Rust
toolchain for the runner crate), `pnpm test:run` and `pnpm build`. CI
runs these.

## Risks

- Low risk. The change adds one `resultJson` key only when the watchdog
moves a run to `failed`, and only when a different earlier code exists.
- Other readers of `resultJson.originalFailureCode`
(`native-run-projection.ts`, `idle-sleep-safety.ts`) only check
committed or succeeded runs. A watchdog failure is neither, so their
results do not change.
- `isPendingNativeRetry` reads only `nativePhase`, so reporting is not
suppressed.
- Sentry groups will split by root cause after deployment. New events
will go to new groups, and the existing group will get fewer events.
This is the intended change.

## Model Used

- Claude Opus 5.5 (`claude-opus-5-5`), with tool use and code execution
in a Paperclip agent sandbox.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-04 11:51:36 -08:00

351 lines
16 KiB
TypeScript

import { notifyNativeEnvironmentSettlement } from "./native-environment-settlement-notifications.js";
import { notifyExternalAgentSettlement } from "../modules/external-agents/index.js";
import { notifyNativeStatusReassessment } from "./native-status-reassessment-notifications.js";
import { notifyNativeFinalizationWork } from "./native-finalization-notifications.js";
import { notifyUnsafeWorkspaceExportRunChange } from "./unsafe-workspace-export-notifications.js";
import { notifyCancelledWorkspacePreservation } from "./cancelled-workspace-preservation-notifications.js";
import { notifyInstructionCollection } from "./instruction-collection-notifications.js";
import { notifyExecutionRunWork } from "./execution-control-notifications.js";
import { randomUUID } from "node:crypto";
import { logger } from "../middleware/logger.js";
import { and, asc, eq, isNotNull, lte, sql } from "drizzle-orm";
import {
agents,
heartbeatRuns,
issues,
nativeRunFinalizations,
type Db,
} from "@paperclipai/db";
import { parseIssueExecutionState } from "./issue-execution-policy.js";
import { issueRecoveryActionService } from "./issue-recovery-actions.js";
import { hasLiveLegacyController } from "./legacy-controller-lease.js";
import { reportRunFailure } from "./run-failure-report.js";
import { getNativeReviewAssignment, readNativeReviewAssignmentContext } from "./native-runtime/native-review-participant.js";
import { preserveNativeFailedWorkspace } from "./native-runtime/native-failed-workspace-recovery.js";
import { EXECUTION_RECONCILIATION_INTERVAL_MS } from "./execution-control-deadline.js";
import { DELIVERY_QUEUES, notifyDeliveryWork } from "./delivery-work-notifications.js";
import { NATIVE_STARTUP_CLEANUP_PENDING, readNativeStartupContext, promoteNativeStartupCleanup } from "./native-runtime/native-startup-cleanup.js";
/** Only newly recorded control deadlines are eligible. Upgrades never replay ambiguous historical runs. */
export async function reconcileAbandonedExecutionControl(
db: Db,
now = new Date(),
options: { startupCleanup?: (scope: { companyId: string; runId: string }) => Promise<unknown> } = {},
) {
const nativeDue = await db
.select({
runId: nativeRunFinalizations.runId,
issueId: nativeRunFinalizations.issueId,
companyId: nativeRunFinalizations.companyId,
failureCode: nativeRunFinalizations.failureCode,
leaseOwner: nativeRunFinalizations.leaseOwner,
controllerGeneration: nativeRunFinalizations.controllerGeneration,
attempt: nativeRunFinalizations.attempt,
})
.from(nativeRunFinalizations)
.where(
and(
isNotNull(nativeRunFinalizations.controlDeadlineAt),
lte(nativeRunFinalizations.controlDeadlineAt, now),
),
)
.orderBy(asc(nativeRunFinalizations.controlDeadlineAt), asc(nativeRunFinalizations.runId))
.limit(50);
const controlDue = await db
.select({
runId: heartbeatRuns.id,
companyId: heartbeatRuns.companyId,
context: heartbeatRuns.contextSnapshot,
nativeIssueId: nativeRunFinalizations.issueId,
failureCode: nativeRunFinalizations.failureCode,
leaseOwner: nativeRunFinalizations.leaseOwner,
controllerGeneration: nativeRunFinalizations.controllerGeneration,
attempt: nativeRunFinalizations.attempt,
})
.from(heartbeatRuns)
.leftJoin(nativeRunFinalizations, eq(nativeRunFinalizations.runId, heartbeatRuns.id))
.where(
and(
isNotNull(heartbeatRuns.executionControlDeadlineAt),
lte(heartbeatRuns.executionControlDeadlineAt, now),
),
)
.orderBy(asc(heartbeatRuns.executionControlDeadlineAt), asc(heartbeatRuns.id))
.limit(50);
const due = [
...new Map(
[
...nativeDue,
...controlDue.map((row) => ({
...row,
issueId: row.nativeIssueId ?? (
typeof row.context?.issueId === "string"
? row.context.issueId
: null),
})),
].map((row) => [row.runId, row]),
).values(),
];
let surfaced = 0;
// Bound contention latency across independent tasks: a locked task must not
// consume the entire reconciliation window for every task behind it.
let nextCandidate = 0;
await Promise.all(Array.from({ length: Math.min(5, due.length) }, async () => {
while (nextCandidate < due.length) {
const candidate = due[nextCandidate++]!;
try {
if (candidate.failureCode === NATIVE_STARTUP_CLEANUP_PENDING) {
if (!options.startupCleanup) continue;
try {
await options.startupCleanup({ companyId: candidate.companyId, runId: candidate.runId });
} finally {
await db.transaction(async tx => {
if (candidate.issueId) await tx.select({ id: issues.id }).from(issues).where(eq(issues.id, candidate.issueId)).for("update");
const [owner] = await tx.select().from(nativeRunFinalizations).where(eq(nativeRunFinalizations.runId, candidate.runId)).for("update");
if (owner?.failureCode !== NATIVE_STARTUP_CLEANUP_PENDING || !owner.controlDeadlineAt || owner.controlDeadlineAt > now ||
owner.leaseOwner !== candidate.leaseOwner || owner.controllerGeneration !== candidate.controllerGeneration || owner.attempt !== candidate.attempt) return;
const cleanupRetryAt = new Date(Math.max(now.getTime(), Date.now()) + EXECUTION_RECONCILIATION_INTERVAL_MS);
await notifyDeliveryWork(tx, DELIVERY_QUEUES.executionFinalization);
await tx.update(nativeRunFinalizations).set({ controlDeadlineAt: cleanupRetryAt }).where(eq(nativeRunFinalizations.runId, candidate.runId));
await tx.update(heartbeatRuns).set({ executionControlDeadlineAt: cleanupRetryAt }).where(and(eq(heartbeatRuns.id, candidate.runId),
isNotNull(heartbeatRuns.executionControlDeadlineAt), lte(heartbeatRuns.executionControlDeadlineAt, now)));
});
}
continue;
}
// Set inside the transaction only when the write below genuinely
// transitions the run into "failed". Read after the transaction
// commits, so a rolled-back write never reports a false failure.
let terminalRunToReport: typeof heartbeatRuns.$inferSelect | null = null;
const repaired = await db.transaction(async (tx) => {
await tx.execute(
sql`select set_config('statement_timeout', '15000', true), set_config('lock_timeout', '1000', true)`,
);
const task = candidate.issueId
? (
await tx
.select()
.from(issues)
.where(
and(
eq(issues.id, candidate.issueId),
eq(issues.companyId, candidate.companyId),
),
)
.for("update")
)[0]
: null;
const [coordinator] = await tx
.select()
.from(nativeRunFinalizations)
.where(
and(
eq(nativeRunFinalizations.runId, candidate.runId),
eq(nativeRunFinalizations.companyId, candidate.companyId),
),
)
.for("update");
const [run] = await tx
.select()
.from(heartbeatRuns)
.where(
and(
eq(heartbeatRuns.id, candidate.runId),
eq(heartbeatRuns.companyId, candidate.companyId),
),
)
.for("update");
if (
!run ||
![
coordinator?.controlDeadlineAt,
run.executionControlDeadlineAt,
].some((deadline) => deadline && deadline <= now)
)
return false;
if (coordinator?.failureCode === NATIVE_STARTUP_CLEANUP_PENDING) return false;
const startupContext = coordinator && readNativeStartupContext(run, coordinator);
if (startupContext) {
if (!await promoteNativeStartupCleanup(tx as unknown as Db, run, coordinator!, now)) {
const retryAt = new Date(Math.max(now.getTime() + EXECUTION_RECONCILIATION_INTERVAL_MS,
startupContext.deadlineAtMs, coordinator!.leaseExpiresAt?.getTime() ?? 0));
await notifyDeliveryWork(tx, DELIVERY_QUEUES.executionFinalization);
await tx.update(nativeRunFinalizations).set({ controlDeadlineAt: retryAt }).where(eq(nativeRunFinalizations.runId, run.id));
await tx.update(heartbeatRuns).set({ executionControlDeadlineAt: retryAt }).where(eq(heartbeatRuns.id, run.id));
}
return false;
}
await tx
.update(heartbeatRuns)
.set({ executionControlDeadlineAt: null })
.where(eq(heartbeatRuns.id, run.id));
if (
[
"succeeded",
"failed",
"cancelled",
"timed_out",
"interrupted",
].includes(run.status)
) {
if (coordinator?.controlDeadlineAt)
await tx
.update(nativeRunFinalizations)
.set({ controlDeadlineAt: null })
.where(eq(nativeRunFinalizations.runId, run.id));
return true;
}
// A persisted result belongs to the existing finalizer, not a replacement provider.
if (coordinator?.resultId) {
await notifyNativeFinalizationWork(tx, { companyId: run.companyId, runId: run.id });
await tx
.update(nativeRunFinalizations)
.set({
controlDeadlineAt: null,
leaseOwner: null,
leaseExpiresAt: null,
updatedAt: now,
})
.where(eq(nativeRunFinalizations.runId, run.id));
return true;
}
if (await hasLiveLegacyController(tx as unknown as Db, run)) {
await tx
.update(heartbeatRuns)
.set({ executionControlDeadlineAt: run.controllerLeaseExpiresAt })
.where(eq(heartbeatRuns.id, run.id));
return false;
}
const cause = "execution_finalization_deadline_exceeded";
const nextAction =
"Inspect the failed run and verify its provider has stopped. Reconcile any uncertain external action before explicitly continuing this task.";
// The deadline replaces the run's error code. Keep the earlier
// diagnosis as its root cause, as the native finalizer does.
const priorFailureCode = coordinator?.failureCode ?? run.errorCode;
const recordsOriginalFailure = Boolean(priorFailureCode) && priorFailureCode !== cause &&
typeof run.resultJson?.originalFailureCode !== "string";
if (coordinator)
await tx
.update(nativeRunFinalizations)
.set({
phase: "terminal_failure",
controlDeadlineAt: null,
leaseOwner: null,
leaseExpiresAt: null,
recoveryState: "blocked",
nextAttemptAt: null,
failureCode: cause,
failureDetail: {
...coordinator.failureDetail,
nextAction,
recoveryOwner: { kind: "board" },
},
updatedAt: now,
})
.where(eq(nativeRunFinalizations.runId, run.id));
await notifyInstructionCollection(tx, { companyId: run.companyId, runId: run.id });
await notifyExternalAgentSettlement(tx, { companyId: run.companyId, runId: run.id });
await notifyNativeEnvironmentSettlement(tx, { companyId: run.companyId, runId: run.id });
await notifyCancelledWorkspacePreservation(tx, { companyId: run.companyId, runId: run.id });
await notifyUnsafeWorkspaceExportRunChange(tx, run);
await notifyExecutionRunWork(tx, { companyId: run.companyId });
const [updatedRun] = await tx
.update(heartbeatRuns)
.set({
status: "failed",
executionStatusDeliveryId: randomUUID(),
finishedAt: now,
...(coordinator
? { nativePhase: "terminal_failure", nativePhaseUpdatedAt: now }
: {}),
errorCode: cause,
error: nextAction,
nextAction,
...(recordsOriginalFailure
? { resultJson: { ...run.resultJson, originalFailureCode: priorFailureCode } }
: {}),
updatedAt: now,
})
.where(eq(heartbeatRuns.id, run.id))
.returning();
if (updatedRun && updatedRun.status !== run.status) {
terminalRunToReport = updatedRun;
}
await tx
.update(agents)
.set({ status: "idle", updatedAt: now })
.where(
and(
eq(agents.id, run.agentId),
eq(agents.companyId, run.companyId),
eq(agents.status, "running"),
eq(agents.lifecycleState, "ready"),
sql`not exists (select 1 from ${heartbeatRuns} where ${heartbeatRuns.agentId} = ${run.agentId} and ${heartbeatRuns.status} = 'running')`,
),
);
const nativeReviewContext = readNativeReviewAssignmentContext(run.contextSnapshot);
const nativeReviewAssignment = nativeReviewContext && task
? await getNativeReviewAssignment(tx as unknown as Db, {
companyId: run.companyId,
issueId: task.id,
agentId: run.agentId,
contextSnapshot: nativeReviewContext,
})
: null;
const review = task?.status === "in_review" ? parseIssueExecutionState(task.executionState) : null;
const isCurrentReviewer = (review?.status === "pending" &&
review.currentParticipant?.type === "agent" && review.currentParticipant.agentId === run.agentId)
|| nativeReviewAssignment?.interaction.status === "pending";
if (
!task ||
(task.assigneeAgentId !== run.agentId && !isCurrentReviewer) ||
["done", "cancelled"].includes(task.status) ||
(task.executionRunId && task.executionRunId !== run.id) ||
(task.checkoutRunId && task.checkoutRunId !== run.id)
) {
if (updatedRun) await preserveNativeFailedWorkspace(tx as unknown as Db, updatedRun);
return true;
}
await notifyNativeStatusReassessment(tx, { companyId: task.companyId, issueId: task.id });
await tx
.update(issues)
.set({ executionRunId: null, checkoutRunId: null, updatedAt: now })
.where(eq(issues.id, task.id));
await issueRecoveryActionService(
tx as unknown as Db,
).upsertSourceScoped({
companyId: run.companyId,
sourceIssueId: task.id,
kind: "active_run_watchdog",
ownerType: "board",
ownerAgentId: null,
returnOwnerAgentId: task.assigneeAgentId,
cause,
fingerprint: `execution-control:${run.id}`,
evidence: {
runId: run.id,
...(isCurrentReviewer ? { reviewParticipantAgentId: run.agentId } : {}),
originalFailureCode: coordinator?.failureCode ?? run.errorCode,
providerOwnership: "unverified",
},
nextAction,
wakePolicy: null,
maxAttempts: 3,
supersedeOnIdentityChange: true,
});
if (updatedRun) await preserveNativeFailedWorkspace(tx as unknown as Db, updatedRun);
return true;
});
if (repaired) surfaced += 1;
if (terminalRunToReport) void reportRunFailure(db, terminalRunToReport);
} catch {
logger.warn(
{ runId: candidate.runId },
"Execution finalization reconciliation remains pending; continuing with other runs",
);
}
}
}));
return { scanned: due.length, surfaced };
}