Files
PaperClipAI/server/src/services/native-runtime/native-run-finalizer-telemetry.test.ts
T
Dotta 1dceee9a4e fix(runner): persist warm Daytona workspaces (#12901)
## Thinking Path

> - Paperclip manages AI agent work and the execution state for each
task.
> - Remote agents run in sandbox environments such as Daytona.
> - Daytona keeps files while a sandbox is stopped, but deletion removes
those files.
> - Runner Codex did not copy successful remote workspace changes back
to the host workspace.
> - A warm sandbox could therefore hide data loss until Daytona replaced
or deleted the sandbox.
> - This pull request makes the host workspace durable after every
successful turn and keeps verified reusable sandboxes warm.
> - The benefit is reliable multi-turn work across warm reuse, restart,
stop, and sandbox replacement.

## Linked Issues or Issue Description

**What happened?**

A successful native Codex turn in Daytona could leave workspace changes
only in the remote sandbox. A later warm turn appeared to work because
it reused that filesystem. A replacement sandbox could start from stale
host data and lose the successful changes.

**Expected behavior**

Paperclip must merge each successful remote turn into the authoritative
host workspace before it completes the run. A verified warm lease may
reuse its remote files. A replacement lease must reconstruct the exact
durable workspace seed.

**Steps to reproduce**

1. Run Codex in a reusable Daytona environment.
2. Write a file during one successful turn.
3. Replace the Daytona sandbox before the next turn.
4. Observe that the next turn can start without the prior file on the
unpatched code.

Related remote workspace foundation: #10070.

## What Changed

- Added explicit `host_current`, `durable_seed`, and `adopt_remote`
workspace preparation modes.
- Added atomic, versioned native workspace descriptors and seed archives
under `PAPERCLIP_HOME`.
- Added real native sandbox export and three-way host merge before
terminal result completion.
- Added workspace-only recovery after a proposed result. Recovery does
not submit another provider turn or consume the provider retry budget.
- Added fail-closed handling when a sandbox with unexported changes is
gone.
- Kept healthy reusable Daytona sandboxes started for legacy Codex and
Runner Codex.
- Kept the Runner Codex process and provider session across verified
warm turns.
- Added the paid `daytona-warm-continuity` browser suite. It contains
exactly the legacy Codex and Runner Codex cells. Each cell performs
three measured turns.
- Documented `pnpm test:e2e:runner -- --suite daytona-warm-continuity`.
No package script was added.
- Added no database migration. The metadata format is backward
compatible and idempotent.

## Verification

- `pnpm typecheck`
- `pnpm test:e2e:runner:unit` — 114 passed
- Native workspace, finalizer, session, and environment tests — 232
passed
- Daytona provider tests — 150 passed
- Workspace staging and merge tests — 98 passed
- Runner transport tests — 63 passed
- Legacy Codex restore tests — 5 passed
- Rust format and compile checks pass through root typecheck
- The paid Daytona suite was not run locally because the required
Daytona, OpenAI, and immutable image credentials are not present.

## Risks

- The main risk is an incorrect workspace identity or merge after a
crash. Durable descriptors bind the run, workspace, lease, provider
lease, local root, remote root, and baseline digest. Ambiguous evidence
fails closed.
- The host merge may conflict with concurrent host edits. The existing
three-way merge and exclusion rules handle this case and surface
failures.
- A deleted sandbox cannot recover unexported bytes. Paperclip now
blocks with `workspace_sync_out_unrecoverable` instead of reporting
success or rerunning the provider.
- There is no database migration. Descriptor writes and recovery are
atomic and idempotent.

## Model Used

OpenAI Codex with GPT-5. The run used agentic reasoning, repository
inspection, code execution, test execution, Git, and GitHub CLI tools.

## 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
- [x] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge
2026-09-05 13:00:57 -05:00

463 lines
16 KiB
TypeScript

import { randomUUID } from "node:crypto";
import { eq } from "drizzle-orm";
import { afterAll, beforeAll, describe, expect, it, vi } from "vitest";
import {
agents,
companies,
completionContracts,
createDb,
heartbeatRuns,
issues,
nativeRunFinalizations,
nativeRunResults,
workAssessments,
} from "@paperclipai/db";
import {
getEmbeddedPostgresTestSupport,
startEmbeddedPostgresTestDatabase,
} from "../../__tests__/helpers/embedded-postgres.js";
import {
CONTROL_PLANE_CONFORMANCE_RESULT,
CONTROL_PLANE_CONFORMANCE_TERMINAL,
} from "../../vendor/paperclip-runner/testing.js";
const mockTelemetryClient = vi.hoisted(() => ({
track: vi.fn(),
hashPrivateRef: vi.fn((value: string) => `hashed:${value}`),
}));
vi.mock("../../telemetry.ts", () => ({ getTelemetryClient: () => mockTelemetryClient }));
function agentTaskRunCalls(fromIndex: number) {
return mockTelemetryClient.track.mock.calls
.slice(fromIndex)
.filter((call) => call[0] === "agent.task_run");
}
import { PaperclipControlPlanePort } from "./paperclip-control-plane-port.js";
import { finalizeNativeRun, recordNativeFinalizationFailure } from "./native-run-finalizer.js";
import { commitNativeStatusDecision } from "./status-decision-committer.js";
import { NATIVE_STATUS_ARBITER_POLICY_VERSION, type NativeStatusDecision } from "./status-arbiter.js";
const embeddedPostgresSupport = await getEmbeddedPostgresTestSupport();
const describeEmbeddedPostgres = embeddedPostgresSupport.supported ? describe : describe.skip;
if (!embeddedPostgresSupport.supported) {
console.warn(
`Skipping embedded Postgres native-finalizer telemetry tests on this host: ${
embeddedPostgresSupport.reason ?? "unsupported environment"
}`,
);
}
describeEmbeddedPostgres("native run finalizer / status decision committer — agent.task_run emission", () => {
let db!: ReturnType<typeof createDb>;
let tempDb: Awaited<ReturnType<typeof startEmbeddedPostgresTestDatabase>> | null = null;
let companyId: string;
let agentId: string;
beforeAll(async () => {
tempDb = await startEmbeddedPostgresTestDatabase("paperclip-native-finalizer-telemetry-");
db = createDb(tempDb.connectionString);
companyId = randomUUID();
agentId = randomUUID();
await db.insert(companies).values({ id: companyId, name: "Telemetry", issuePrefix: "TLM" });
await db.insert(agents).values({
id: agentId,
companyId,
name: "Native worker",
adapterType: "codex_local",
status: "running",
});
}, 30_000);
afterAll(async () => {
await tempDb?.cleanup();
});
async function seedNativeRun() {
const issueId = randomUUID();
const contractId = randomUUID();
const runId = randomUUID();
const sessionId = randomUUID();
const runnerInstanceId = randomUUID();
const contractSha256 = `contract-${contractId}`;
await db.insert(issues).values({
id: issueId,
companyId,
title: "Native telemetry fixture",
status: "in_progress",
assigneeAgentId: agentId,
workMode: "standard",
});
await db.insert(completionContracts).values({
id: contractId,
companyId,
issueId,
revision: 1,
schemaVersion: "paperclip.completion-contract.v1",
policyVersion: "telemetry-v1",
risk: "standard",
completionAuthority: "server_arbiter",
incompleteCriteriaPolicy: "preserve_non_terminal",
contractJson: {
revision: "telemetry-v1",
objective: "Emit telemetry",
criteria: [{ id: "objective", requirement: "Complete the task" }],
},
canonicalSha256: contractSha256,
createdByActorType: "system",
createdByActorId: "test",
});
await db.insert(heartbeatRuns).values({
id: runId,
companyId,
agentId,
status: "running",
runtimeMode: "native",
nativeIssueId: issueId,
nativeSessionId: sessionId,
runnerInstanceId,
completionContractId: contractId,
completionContractSha256: contractSha256,
contextSnapshot: { issueId },
});
return { issueId, contractId, contractSha256, runId, sessionId, runnerInstanceId };
}
function newPort(fixture: Awaited<ReturnType<typeof seedNativeRun>>) {
return new PaperclipControlPlanePort(db, {
companyId,
issueId: fixture.issueId,
runId: fixture.runId,
agentId,
sessionId: fixture.sessionId,
completionContractId: fixture.contractId,
completionContractSha256: fixture.contractSha256,
sourceInstanceId: fixture.runnerInstanceId,
controlPlaneSourceInstanceId: "telemetry-test",
});
}
async function driveToCompleteResult(
fixture: Awaited<ReturnType<typeof seedNativeRun>>,
terminal: typeof CONTROL_PLANE_CONFORMANCE_TERMINAL = CONTROL_PLANE_CONFORMANCE_TERMINAL,
) {
const port = newPort(fixture);
await port.openRun({
identity: {
runId: fixture.runId,
sessionId: fixture.sessionId,
companyId,
issueId: fixture.issueId,
agentId,
},
backendKind: "mock",
sourceInstanceId: fixture.runnerInstanceId,
});
await port.completeRun({
result: CONTROL_PLANE_CONFORMANCE_RESULT,
terminal,
callerResultId: `${fixture.runId}:result`,
});
}
it("emits exactly one agent.task_run event when the native finalizer commits a terminal status", async () => {
const fixture = await seedNativeRun();
await driveToCompleteResult(fixture);
const callsBefore = mockTelemetryClient.track.mock.calls.length;
await finalizeNativeRun({
db,
runId: fixture.runId,
workspaceFinalizeStatus: "succeeded",
projectRunStatus: true,
});
const newCalls = agentTaskRunCalls(callsBefore);
expect(newCalls).toHaveLength(1);
expect(newCalls[0]?.[1]).toMatchObject({ agent_id: agentId, state: "succeeded" });
const run = await db
.select({ status: heartbeatRuns.status })
.from(heartbeatRuns)
.where(eq(heartbeatRuns.id, fixture.runId))
.then((rows) => rows[0]);
expect(run?.status).toBe("succeeded");
});
it("emits zero events when a repeat finalize call's conditional native write matches no row", async () => {
const fixture = await seedNativeRun();
await driveToCompleteResult(fixture);
await finalizeNativeRun({
db,
runId: fixture.runId,
workspaceFinalizeStatus: "succeeded",
projectRunStatus: true,
});
const callsBefore = mockTelemetryClient.track.mock.calls.length;
// The coordinator is already "committed", so this second call takes the
// projectCommittedRun short-circuit. The run's status is already terminal
// ("succeeded"), which sits outside projectCommittedRun's WHERE
// (queued/running/failed), so the write matches no row.
await finalizeNativeRun({
db,
runId: fixture.runId,
workspaceFinalizeStatus: "succeeded",
projectRunStatus: true,
});
expect(agentTaskRunCalls(callsBefore)).toHaveLength(0);
});
it("emits zero events when a reconciliation replay commits the same failed terminal result again", async () => {
const fixture = await seedNativeRun();
await driveToCompleteResult(fixture, {
...CONTROL_PLANE_CONFORMANCE_TERMINAL,
turnTerminalState: "failed",
runTerminalState: "failed",
});
// workspaceFinalizeStatus reports whether the workspace finalization
// step itself succeeded, independent of the run's own terminal state
// (runTerminalState below), which is what actually failed here.
await finalizeNativeRun({
db,
runId: fixture.runId,
workspaceFinalizeStatus: "succeeded",
projectRunStatus: true,
});
const run = await db
.select({ status: heartbeatRuns.status })
.from(heartbeatRuns)
.where(eq(heartbeatRuns.id, fixture.runId))
.then((rows) => rows[0]);
expect(run?.status).toBe("failed");
const callsBefore = mockTelemetryClient.track.mock.calls.length;
// The coordinator is already "committed" with a "failed" terminal
// result, and the run row is already "failed" — unlike the "succeeded"
// case above, "failed" sits INSIDE projectCommittedRun's WHERE clause
// (queued/running/failed), so a reconciliation replay's write still
// matches the row. The write changes nothing (failed -> failed), so it
// must not emit a second event for the same committed result.
await finalizeNativeRun({
db,
runId: fixture.runId,
workspaceFinalizeStatus: "succeeded",
projectRunStatus: true,
});
expect(agentTaskRunCalls(callsBefore)).toHaveLength(0);
});
it("emits zero events when a retryable-failure write's conditional status spread is omitted", async () => {
const fixture = await seedNativeRun();
// recordNativeFinalizationFailure only needs the run and coordinator rows
// to exist — no persisted native result is required.
await db.insert(nativeRunFinalizations).values({
runId: fixture.runId,
companyId,
issueId: fixture.issueId,
phase: "observed",
attempt: 0,
});
const callsBefore = mockTelemetryClient.track.mock.calls.length;
// projectRunStatus is omitted (falsy): the conditional spread in
// recordRetryableFailure never includes `status`, so the write never sets
// a terminal status, regardless of the attempt count.
await recordNativeFinalizationFailure({
db,
runId: fixture.runId,
error: new Error("native_test_failure"),
});
expect(agentTaskRunCalls(callsBefore)).toHaveLength(0);
const run = await db
.select({ status: heartbeatRuns.status })
.from(heartbeatRuns)
.where(eq(heartbeatRuns.id, fixture.runId))
.then((rows) => rows[0]);
expect(run?.status).toBe("running");
});
it("bounds workspace-only retries without consuming the provider attempt", async () => {
const fixture = await seedNativeRun();
await db.insert(nativeRunFinalizations).values({
runId: fixture.runId,
companyId,
issueId: fixture.issueId,
phase: "observed",
attempt: 1,
});
for (
let expectedAttempt = 1;
expectedAttempt <= 3;
expectedAttempt += 1
) {
await recordNativeFinalizationFailure({
db,
runId: fixture.runId,
error: new Error("native_workspace_sync_out_failed"),
projectRunStatus: true,
failureScope: "workspace",
});
const coordinator = await db
.select()
.from(nativeRunFinalizations)
.where(eq(nativeRunFinalizations.runId, fixture.runId))
.then((rows) => rows[0]!);
expect(coordinator.attempt).toBe(1);
expect(coordinator.failureDetail).toMatchObject({
workspaceFinalizeAttempt: expectedAttempt,
});
expect(coordinator.phase).toBe(
expectedAttempt === 3 ? "terminal_failure" : "retryable_failure",
);
}
await expect(
db
.select({ status: heartbeatRuns.status })
.from(heartbeatRuns)
.where(eq(heartbeatRuns.id, fixture.runId))
.then((rows) => rows[0]?.status),
).resolves.toBe("failed");
});
it("blocks immediately when the sandbox with unexported changes is gone", async () => {
const fixture = await seedNativeRun();
await db.insert(nativeRunFinalizations).values({
runId: fixture.runId,
companyId,
issueId: fixture.issueId,
phase: "workspace_finalizing",
attempt: 1,
resultId: null,
});
const failure = await recordNativeFinalizationFailure({
db,
runId: fixture.runId,
error: new Error("native_workspace_sync_out_unrecoverable"),
projectRunStatus: true,
failureScope: "workspace",
permanent: true,
});
expect(failure).toMatchObject({
phase: "terminal_failure",
failureCode: "native_workspace_sync_out_unrecoverable",
nextAttemptAt: null,
attempt: 1,
});
await expect(
db
.select({ status: heartbeatRuns.status })
.from(heartbeatRuns)
.where(eq(heartbeatRuns.id, fixture.runId))
.then((rows) => rows[0]?.status),
).resolves.toBe("failed");
await expect(
db
.select({ status: issues.status })
.from(issues)
.where(eq(issues.id, fixture.issueId))
.then((rows) => rows[0]?.status),
).resolves.toBe("blocked");
});
it("emits exactly one event for a cancel_continuations write (trap 2: :685/:518 overlap)", async () => {
const fixture = await seedNativeRun();
// Build the minimal real rows commitNativeStatusDecision's foreign keys
// require, without going through the evidence-classifier pipeline.
const resultId = randomUUID();
await db.insert(nativeRunResults).values({
id: resultId,
companyId,
issueId: fixture.issueId,
runId: fixture.runId,
completionContractId: fixture.contractId,
serverFingerprint: `fp-${resultId}`,
schemaStatus: "accepted",
resultJson: {},
canonicalSha256: `sha-${resultId}`,
});
const assessmentId = randomUUID();
await db.insert(workAssessments).values({
id: assessmentId,
companyId,
issueId: fixture.issueId,
runId: fixture.runId,
contractId: fixture.contractId,
resultId,
triggerKind: "native_result",
triggerActorCompanyId: companyId,
priorIssueStatus: "in_progress",
priorStatusVersion: 0,
policyVersion: "telemetry-v1",
assessmentJson: {},
inputDigest: `digest-${assessmentId}`,
});
await db.insert(nativeRunFinalizations).values({
runId: fixture.runId,
companyId,
issueId: fixture.issueId,
phase: "arbitrating",
attempt: 0,
resultId,
});
// Shaped exactly like resolveNativeCancellationStatus("issue", ...) in
// native-session-executor.ts — the only place cancel_continuations is
// produced today.
const decision: NativeStatusDecision = {
policyVersion: NATIVE_STATUS_ARBITER_POLICY_VERSION,
statusAction: "cancelled",
toStatus: "cancelled",
reasonCode: "cancellation_issue_authorized",
unblockDescriptor: null,
effects: [
{ kind: "release_checkout" },
{ kind: "cancel_continuations" },
],
};
const callsBefore = mockTelemetryClient.track.mock.calls.length;
await commitNativeStatusDecision({
db,
companyId,
issueId: fixture.issueId,
runId: fixture.runId,
assessmentId,
priorStatus: "in_progress",
priorStatusVersion: 0,
priorDecisionId: null,
decision,
});
const afterCommit = agentTaskRunCalls(callsBefore);
// status-decision-committer.ts:685 (the cancel_continuations effect)
// wrote this run's terminal status and emitted for it.
expect(afterCommit).toHaveLength(1);
expect(afterCommit[0]?.[1]).toMatchObject({ agent_id: agentId, state: "cancelled" });
const run = await db
.select({ status: heartbeatRuns.status })
.from(heartbeatRuns)
.where(eq(heartbeatRuns.id, fixture.runId))
.then((rows) => rows[0]);
expect(run?.status).toBe("cancelled");
// native-run-finalizer.ts:518's own guard: it skips its emit whenever the
// decision it just committed already included cancel_continuations,
// because :685 above already wrote and emitted for this run. Reproduce
// the exact guard predicate on the same decision object to prove it
// resolves to "skip", so the run's total event count for this commit
// stays at exactly one even after the finalizer's own write runs.
const alreadyEmittedByCommittedDecision = decision.effects.some(
(effect) => effect.kind === "cancel_continuations",
);
expect(alreadyEmittedByCommittedDecision).toBe(true);
expect(agentTaskRunCalls(callsBefore)).toHaveLength(1);
});
});