mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-10-07 16:11:46 +02:00
## Thinking Path > - Paperclip is the control plane for agents that perform work. > - Paperclip Runner connects durable provider sessions to individual task runs through PRP. > - Provider continuity and per-run authority are different lifetimes. > - The existing implementation mixed those lifetimes and lost event metadata between provider frames, runnerd, persistence, API sanitization, and the task thread. > - That caused failed continuation, missing progress and Plans, duplicate replies, hidden failures, and unsafe recovery. > - This repair gives every heartbeat fresh authority, preserves qualified provider-session continuity, and restores one lossless presentation path without changing direct adapters. ## Linked Issues or Issue Description **What happened?** A second native heartbeat could reuse tickets, leases, command receipts, sequence state, and run identity from the first heartbeat. Provider phase and item identity could be lost before the UI read them. Redaction could corrupt protocol discriminators while still missing malformed credential tails. The task thread could fold progress into the final response, hide failures, or show more than one final answer. Native Codex also exposed approval modes that do not yet have a durable approval bridge. **Expected behavior** Each heartbeat uses a new PRP authority epoch. Codex and OpenCode preserve exact qualified provider sessions; ACPX emits an explicit continuity event when its qualified process-replacement policy is used. Every accepted provider event is presented, classified as internal, or surfaced as unsupported. The task page shows chronological progress, reasoning summaries, activity, Plans, interactions, terminal failures, and exactly one final reply. Direct adapters retain their existing path. **Steps to reproduce** 1. Enable the unified experimental Paperclip Runner setting. 2. Create a local native Codex, OpenCode, ACPX Claude, or ACPX Codex agent. 3. Run response, Plan, structured-question/resume, restart, cancellation, and failure scenarios. 4. Reload the task while active, waiting, failed, and settled. 5. On the old implementation, observe stale run authority, missing classifications, incomplete output, or duplicated/folded replies. **Paperclip version or commit** The repair is based directly on `master` at `87d05e194b643810d16d20612115acd01d735d43`. **Deployment mode** Local development with the embedded database. Related work: Refs #12616, #12646, #12666, #12685, and #12700. ## What Changed - Rotates PRP control-plane, outbox, ticket, lease, command, receipt, and sequence authority for each heartbeat while carrying forward only a validated provider-session identity. - Reads `control-plane-state.json`, validates both durable schemas and lifecycle values, resumes coherent current runs, archives qualified settled authority, and quarantines malformed or mismatched scoped state without moving ambiguous live legacy state. - Preserves Codex provider phase and stable item identities so commentary remains progress and only `final_answer` becomes final. - Adds raw OpenCode HTTP/SSE boundary coverage and canonical reasoning lifecycle mapping. - Makes ACPX normalization lossless for visible reasoning, tool lifecycle metadata, stable bounded identities, Plan revisions, structured requests, failures, and qualified process replacement. Only the compatible terminal assistant message is promoted as final. - Applies schema-aware redaction before generic JWT-shaped detection and scans every diagnostic string leaf. Malformed raw/escaped quoted credential tails are redacted in both server and durable Rust state. - Restores snapshot-style chronological task presentation, expandable tool activity, inline Plan cards, visible waiting/resume/cancel/failure states, and exactly one final answer. - Makes `never` the only qualified native Codex permission mode and rejects unsupported persisted native modes with remediation. OpenCode and ACPX policies remain intact. - Keeps the unified experimental Runner setting as the only enablement flag. Onboarding and direct Codex, Claude, and OpenCode stay on their legacy execution/finalization paths. - Adds cross-language goldens, authority/recovery/fault coverage, exact response/count assertions, and native plus legacy acceptance scenarios. ## Verification - Pull-request GitHub Actions run Rust formatting/tests, TypeScript checks, server/UI tests, builds, protocol drift checks, browser E2E, and security scans. - A separate workflow-only validation ref is pinned directly on this PR head and runs the 35-cell paid local matrix: three core scenarios plus structured-question resume and restart/resume for native Codex, native OpenCode, ACPX Claude, ACPX Codex, and direct Codex/Claude/OpenCode. Run: https://github.com/paperclipai/paperclip/actions/runs/33682434315 - Acceptance requires exact single visible replies, monotonic sequences, matching envelope discriminators, one semantic terminal, one run terminal, no unresolved interaction, no duplicate mutation, no secret leakage, provider continuity, and zero native rows for direct adapters. - Per maintainer direction, tests are running in GitHub Actions rather than on the slower local host. Only formatters and static diff checks were run locally. ## Risks - Recovery from old or partial filesystem state is sensitive. The repair fails closed, preserves active or unverifiable authority, and quarantines only state whose scoped ownership is safe to move. - Provider event formats can change. Closed validators and boundary goldens turn new or malformed events into visible diagnostics instead of silent drops. - Shared task presentation could affect direct adapters. Runtime-fact gating plus the direct-adapter matrix protect the existing path. - Managed and remote providers are not qualified here. Shared code continues to compile and fail safely, but live qualification is deferred. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used OpenAI Codex based on GPT-5. The exact deployed snapshot and context-window size are not exposed to this task. It used agentic reasoning, repository inspection, code editing, Git, parallel subagents, and GitHub Actions. ## 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 and contains no internal Paperclip ticket id - [ ] I have run tests locally and they pass (intentionally deferred to GitHub Actions) - [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 risks above - [ ] All Paperclip CI gates are green - [ ] The paid local-provider matrix is 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
393 lines
12 KiB
TypeScript
393 lines
12 KiB
TypeScript
import { spawn, type ChildProcess } from "node:child_process";
|
|
import { createWriteStream } from "node:fs";
|
|
import { mkdir, readFile, rename, writeFile } from "node:fs/promises";
|
|
import path from "node:path";
|
|
import {
|
|
assertIsolatedServerEnvironment,
|
|
buildPaperclipServerEnvironment,
|
|
} from "./harness-env.js";
|
|
|
|
function required(name: string) {
|
|
const value = process.env[name]?.trim();
|
|
if (!value) throw new Error(`${name} is required`);
|
|
return value;
|
|
}
|
|
|
|
const logPath = required("PAPERCLIP_RUNNER_E2E_SERVER_LOG");
|
|
const temporaryRoot = required("PAPERCLIP_RUNNER_E2E_TEMP_ROOT");
|
|
const paperclipHome = required("PAPERCLIP_HOME");
|
|
const configPath = required("PAPERCLIP_CONFIG");
|
|
const port = required("PAPERCLIP_RUNNER_E2E_PORT");
|
|
const repositoryRoot = path.resolve(import.meta.dirname, "../..");
|
|
const tsxCli = path.join(repositoryRoot, "cli/node_modules/tsx/dist/cli.mjs");
|
|
const paperclipCli = path.join(repositoryRoot, "cli/src/index.ts");
|
|
const controlDirectory = path.join(temporaryRoot, "control");
|
|
const restartRequestPath = path.join(
|
|
controlDirectory,
|
|
"server-restart.request.json",
|
|
);
|
|
const restartAckPath = path.join(controlDirectory, "server-restart.ack.json");
|
|
const restartTimeoutMs = 180_000;
|
|
const gracefulStopTimeoutMs = 30_000;
|
|
const serverEnvironment = buildPaperclipServerEnvironment(process.env, {
|
|
NODE_ENV: "test",
|
|
PORT: port,
|
|
PAPERCLIP_HOME: paperclipHome,
|
|
PAPERCLIP_CONFIG: configPath,
|
|
PAPERCLIP_INSTANCE_ID: required("PAPERCLIP_INSTANCE_ID"),
|
|
PAPERCLIP_AGENT_JWT_SECRET: required("PAPERCLIP_AGENT_JWT_SECRET"),
|
|
PAPERCLIP_DECISION_SIGNING_SECRET: required(
|
|
"PAPERCLIP_DECISION_SIGNING_SECRET",
|
|
),
|
|
PAPERCLIP_TOOL_ACTION_SIGNING_SECRET: required(
|
|
"PAPERCLIP_TOOL_ACTION_SIGNING_SECRET",
|
|
),
|
|
BETTER_AUTH_SECRET: required("BETTER_AUTH_SECRET"),
|
|
PAPERCLIP_BIND: "loopback",
|
|
PAPERCLIP_BIND_HOST: "127.0.0.1",
|
|
PAPERCLIP_DEPLOYMENT_MODE: "local_trusted",
|
|
PAPERCLIP_DEPLOYMENT_EXPOSURE: "private",
|
|
SERVE_UI: "true",
|
|
PAPERCLIP_STORAGE_PROVIDER: "local_disk",
|
|
PAPERCLIP_STORAGE_LOCAL_DIR: path.join(temporaryRoot, "storage"),
|
|
PAPERCLIP_SECRETS_PROVIDER: "local_encrypted",
|
|
PAPERCLIP_SECRETS_STRICT_MODE: "true",
|
|
PAPERCLIP_DB_BACKUP_ENABLED: "false",
|
|
PAPERCLIP_DB_BACKUP_DIR: path.join(temporaryRoot, "backups"),
|
|
// Onboarding normally opens the app after listen. Browser ownership belongs
|
|
// to Playwright in this harness, so never create a developer desktop tab.
|
|
PAPERCLIP_OPEN_ON_LISTEN: "false",
|
|
});
|
|
assertIsolatedServerEnvironment(serverEnvironment, {
|
|
temporaryRoot,
|
|
paperclipHome,
|
|
configPath,
|
|
});
|
|
const definedServerEnvironment = Object.fromEntries(
|
|
Object.entries(serverEnvironment).filter(
|
|
(entry): entry is [string, string] => entry[1] !== undefined,
|
|
),
|
|
);
|
|
|
|
await Promise.all([
|
|
mkdir(path.dirname(logPath), { recursive: true }),
|
|
mkdir(controlDirectory, { recursive: true, mode: 0o700 }),
|
|
]);
|
|
const log = createWriteStream(logPath, { flags: "a", mode: 0o600 });
|
|
const expectedStops = new WeakSet<ChildProcess>();
|
|
const childErrors = new WeakMap<ChildProcess, Error>();
|
|
let child: ChildProcess | null = null;
|
|
let unexpectedChildFailure: Error | null = null;
|
|
let shutdownSignal: NodeJS.Signals | null = null;
|
|
let activeRestartRequestId: string | null = null;
|
|
|
|
function appendLog(message: string) {
|
|
process.stderr.write(message);
|
|
log.write(message);
|
|
}
|
|
|
|
function shutdownRequested() {
|
|
return shutdownSignal !== null;
|
|
}
|
|
|
|
function childExited(candidate: ChildProcess) {
|
|
return candidate.exitCode !== null || candidate.signalCode !== null;
|
|
}
|
|
|
|
function describeChildExit(candidate: ChildProcess) {
|
|
const spawnError = childErrors.get(candidate);
|
|
if (spawnError) return `server spawn failed: ${spawnError.message}`;
|
|
return `server exited code=${String(candidate.exitCode)} signal=${String(candidate.signalCode)}`;
|
|
}
|
|
|
|
function startServer() {
|
|
if (shutdownRequested()) {
|
|
throw new Error("Refusing to start Paperclip after wrapper shutdown");
|
|
}
|
|
const candidate = spawn(
|
|
process.execPath,
|
|
[tsxCli, paperclipCli, "onboard", "--yes", "--run"],
|
|
{
|
|
cwd: repositoryRoot,
|
|
env: definedServerEnvironment,
|
|
stdio: ["ignore", "pipe", "pipe"],
|
|
// Stay in the launcher-created process group. That lets the launcher stop
|
|
// Playwright, this wrapper, Paperclip, embedded Postgres, and runner children
|
|
// as one verified tree even if graceful web-server shutdown stalls.
|
|
detached: false,
|
|
},
|
|
);
|
|
child = candidate;
|
|
|
|
candidate.stdout?.on("data", (chunk) => {
|
|
process.stdout.write(chunk);
|
|
log.write(chunk);
|
|
});
|
|
candidate.stderr?.on("data", (chunk) => {
|
|
process.stderr.write(chunk);
|
|
log.write(chunk);
|
|
});
|
|
candidate.once("error", (error) => {
|
|
childErrors.set(candidate, error);
|
|
if (!expectedStops.has(candidate) && !shutdownRequested()) {
|
|
unexpectedChildFailure = new Error(
|
|
`Paperclip server spawn failed: ${error.message}`,
|
|
);
|
|
}
|
|
});
|
|
candidate.once("exit", () => {
|
|
appendLog(`\n${describeChildExit(candidate)}\n`);
|
|
if (!expectedStops.has(candidate) && !shutdownRequested()) {
|
|
unexpectedChildFailure = new Error(
|
|
`Paperclip server stopped unexpectedly: ${describeChildExit(candidate)}`,
|
|
);
|
|
}
|
|
});
|
|
|
|
// A shutdown may arrive in the synchronous interval around spawn. Never let
|
|
// that race create an unowned replacement server.
|
|
if (shutdownSignal) {
|
|
expectedStops.add(candidate);
|
|
try {
|
|
candidate.kill(shutdownSignal);
|
|
} catch {
|
|
// The process may have failed during spawn.
|
|
}
|
|
}
|
|
return candidate;
|
|
}
|
|
|
|
function delay(milliseconds: number) {
|
|
return new Promise<void>((resolve) => setTimeout(resolve, milliseconds));
|
|
}
|
|
|
|
async function waitForExit(candidate: ChildProcess, timeoutMs: number) {
|
|
if (childExited(candidate) || childErrors.has(candidate)) return true;
|
|
return await new Promise<boolean>((resolve) => {
|
|
let settled = false;
|
|
const finish = (exited: boolean) => {
|
|
if (settled) return;
|
|
settled = true;
|
|
clearTimeout(timeout);
|
|
candidate.off("exit", onExit);
|
|
candidate.off("error", onError);
|
|
resolve(exited);
|
|
};
|
|
const onExit = () => finish(true);
|
|
const onError = () => finish(true);
|
|
const timeout = setTimeout(() => finish(false), timeoutMs);
|
|
candidate.once("exit", onExit);
|
|
candidate.once("error", onError);
|
|
});
|
|
}
|
|
|
|
async function stopServer(
|
|
candidate: ChildProcess,
|
|
signal: NodeJS.Signals = "SIGTERM",
|
|
) {
|
|
expectedStops.add(candidate);
|
|
if (childExited(candidate) || childErrors.has(candidate)) return;
|
|
try {
|
|
candidate.kill(signal);
|
|
} catch {
|
|
if (childExited(candidate) || childErrors.has(candidate)) return;
|
|
throw new Error("Could not signal the Paperclip server to stop");
|
|
}
|
|
if (await waitForExit(candidate, gracefulStopTimeoutMs)) return;
|
|
|
|
appendLog(
|
|
`\nPaperclip did not stop within ${gracefulStopTimeoutMs}ms; sending SIGKILL\n`,
|
|
);
|
|
try {
|
|
candidate.kill("SIGKILL");
|
|
} catch {
|
|
if (childExited(candidate) || childErrors.has(candidate)) return;
|
|
throw new Error("Could not force the Paperclip server to stop");
|
|
}
|
|
if (!(await waitForExit(candidate, 5_000))) {
|
|
throw new Error("Paperclip server did not exit after SIGKILL");
|
|
}
|
|
}
|
|
|
|
async function waitForHealth(candidate: ChildProcess) {
|
|
const deadline = Date.now() + restartTimeoutMs;
|
|
const healthUrl = `http://127.0.0.1:${port}/api/health`;
|
|
while (Date.now() < deadline) {
|
|
if (shutdownRequested()) {
|
|
throw new Error("Wrapper shutdown interrupted the Paperclip restart");
|
|
}
|
|
if (childErrors.has(candidate) || childExited(candidate)) {
|
|
throw new Error(
|
|
`Replacement Paperclip server could not start: ${describeChildExit(candidate)}`,
|
|
);
|
|
}
|
|
try {
|
|
const response = await fetch(healthUrl, {
|
|
signal: AbortSignal.timeout(1_000),
|
|
});
|
|
if (response.ok) return;
|
|
} catch {
|
|
// The replacement process may still be booting.
|
|
}
|
|
await delay(250);
|
|
}
|
|
throw new Error(
|
|
`Replacement Paperclip server did not become healthy within ${restartTimeoutMs}ms`,
|
|
);
|
|
}
|
|
|
|
async function waitForHealthToStop() {
|
|
const healthUrl = `http://127.0.0.1:${port}/api/health`;
|
|
const deadline = Date.now() + gracefulStopTimeoutMs;
|
|
while (Date.now() < deadline) {
|
|
if (shutdownRequested()) {
|
|
throw new Error("Wrapper shutdown interrupted the Paperclip restart");
|
|
}
|
|
try {
|
|
await fetch(healthUrl, { signal: AbortSignal.timeout(500) });
|
|
} catch {
|
|
return;
|
|
}
|
|
await delay(100);
|
|
}
|
|
throw new Error(
|
|
"The old Paperclip server remained healthy after its launcher exited",
|
|
);
|
|
}
|
|
|
|
interface RestartRequest {
|
|
requestId: string;
|
|
}
|
|
|
|
async function readRestartRequest(): Promise<RestartRequest | null> {
|
|
let encoded: string;
|
|
try {
|
|
encoded = await readFile(restartRequestPath, "utf8");
|
|
} catch (error) {
|
|
if ((error as NodeJS.ErrnoException).code === "ENOENT") return null;
|
|
throw error;
|
|
}
|
|
let value: unknown;
|
|
try {
|
|
value = JSON.parse(encoded);
|
|
} catch {
|
|
// The writer may not have completed its atomic replacement yet.
|
|
return null;
|
|
}
|
|
if (!value || typeof value !== "object" || Array.isArray(value)) return null;
|
|
const requestId = (value as { requestId?: unknown }).requestId;
|
|
if (
|
|
typeof requestId !== "string" ||
|
|
!/^[A-Za-z0-9._:-]{1,200}$/.test(requestId)
|
|
) {
|
|
return null;
|
|
}
|
|
return { requestId };
|
|
}
|
|
|
|
async function writeRestartAck(
|
|
requestId: string,
|
|
status: "ready" | "failed",
|
|
message?: string,
|
|
) {
|
|
const temporaryAckPath = `${restartAckPath}.${process.pid}.tmp`;
|
|
await writeFile(
|
|
temporaryAckPath,
|
|
`${JSON.stringify({
|
|
requestId,
|
|
status,
|
|
completedAt: new Date().toISOString(),
|
|
...(message ? { message } : {}),
|
|
})}\n`,
|
|
{ encoding: "utf8", mode: 0o600 },
|
|
);
|
|
await rename(temporaryAckPath, restartAckPath);
|
|
}
|
|
|
|
async function restartServer(requestId: string) {
|
|
activeRestartRequestId = requestId;
|
|
appendLog(`\nRestart request ${requestId}: stopping Paperclip\n`);
|
|
const previous = child;
|
|
if (!previous) throw new Error("No Paperclip server is available to restart");
|
|
await stopServer(previous);
|
|
if (child === previous) child = null;
|
|
// Do not mistake an orphaned old server for a healthy replacement. The port
|
|
// must stop answering before the next launcher is allowed to start.
|
|
await waitForHealthToStop();
|
|
if (shutdownRequested()) {
|
|
throw new Error("Wrapper shutdown interrupted the Paperclip restart");
|
|
}
|
|
|
|
appendLog(`Restart request ${requestId}: starting Paperclip\n`);
|
|
const replacement = startServer();
|
|
await waitForHealth(replacement);
|
|
if (shutdownRequested()) {
|
|
throw new Error("Wrapper shutdown interrupted the Paperclip restart");
|
|
}
|
|
await writeRestartAck(requestId, "ready");
|
|
appendLog(`Restart request ${requestId}: Paperclip is healthy\n`);
|
|
activeRestartRequestId = null;
|
|
}
|
|
|
|
for (const signal of ["SIGINT", "SIGTERM", "SIGHUP"] as const) {
|
|
process.on(signal, () => {
|
|
if (shutdownSignal) return;
|
|
shutdownSignal = signal;
|
|
if (!child) return;
|
|
expectedStops.add(child);
|
|
try {
|
|
child.kill(signal);
|
|
} catch {
|
|
// The Paperclip process may already have exited.
|
|
}
|
|
});
|
|
}
|
|
|
|
async function supervise() {
|
|
startServer();
|
|
let lastRestartRequestId: string | null = null;
|
|
while (!shutdownRequested()) {
|
|
if (unexpectedChildFailure) throw unexpectedChildFailure;
|
|
const request = await readRestartRequest();
|
|
if (request && request.requestId !== lastRestartRequestId) {
|
|
lastRestartRequestId = request.requestId;
|
|
await restartServer(request.requestId);
|
|
}
|
|
await delay(200);
|
|
}
|
|
|
|
const running = child;
|
|
if (running) await stopServer(running, shutdownSignal ?? "SIGTERM");
|
|
}
|
|
|
|
let exitCode = 0;
|
|
try {
|
|
await supervise();
|
|
} catch (error) {
|
|
exitCode = 1;
|
|
const message = error instanceof Error ? error.message : String(error);
|
|
appendLog(`\nPaperclip E2E server supervisor failed: ${message}\n`);
|
|
if (activeRestartRequestId) {
|
|
try {
|
|
await writeRestartAck(activeRestartRequestId, "failed", message);
|
|
} catch (ackError) {
|
|
appendLog(
|
|
`Failed to write restart acknowledgement: ${ackError instanceof Error ? ackError.message : String(ackError)}\n`,
|
|
);
|
|
}
|
|
}
|
|
const running = child;
|
|
if (running) {
|
|
try {
|
|
await stopServer(running);
|
|
} catch (stopError) {
|
|
appendLog(
|
|
`Failed to stop Paperclip after supervisor failure: ${stopError instanceof Error ? stopError.message : String(stopError)}\n`,
|
|
);
|
|
}
|
|
}
|
|
}
|
|
|
|
await new Promise<void>((resolve) => log.end(resolve));
|
|
process.exitCode = exitCode;
|