Files
paperclip/scripts/ci/retry-reporter.test.mjs
T
Nicky LeachandClaude Opus 5.5 fb3cf1a072 feat(ci): retry a failed Vitest case once in PR CI and report each retried pass (#19637)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - Contributors wait for PR CI before merge, so CI wall-clock time sets
the pace of work
> - PR CI runs Vitest shards with 4 workers on 4-vCPU runners, so
timing-sensitive tests sometimes time out under load
> - One such timeout fails the whole run, and a re-run costs another 15
minutes or more
> - In one 7-hour window on 2026-11-06, about 30 failed jobs were test
flakes on commits that passed on a later attempt
> - This pull request retries a failed Vitest case once in PR CI only,
and reports every pass that needed the retry
> - The benefit is fewer whole-run re-runs, while each flaky test stays
visible so that we can fix it

## Linked Issues or Issue Description

**What existing behavior does this improve?**

PR CI fails the whole run when one Vitest case fails once because of
runner load.

**Subsystem affected**

CI: the PR Vitest lanes (general, serialized, workspace and Runner
package lanes).

**Current behavior**

Vitest uses `retry: 0` everywhere. One timeout under load fails the run.
A person must notice it and re-run the failed jobs. About 20% of
recently merged PRs needed at least one re-run.

**Proposed behavior**

PR CI retries a failed Vitest case one time. A test that passes only on
the retry gets a GitHub warning annotation and a row in the job summary.
Local runs, release, cloud readiness and master keep `retry: 0`.

**Reason and benefit**

A re-run of one test case costs seconds. A re-run of the CI run costs 15
minutes or more. The report keeps flaky tests visible. Separate pull
requests fix the known flaky tests.

**Breaking changes**

None.

## What Changed

- Add `scripts/ci/vitest-retry.mjs`. It sets 1 retry when
`GITHUB_ACTIONS=true` and `GITHUB_WORKFLOW=PR`, and 0 in other contexts.
`PAPERCLIP_VITEST_RETRY` overrides the value.
- Add `scripts/ci/retry-reporter.mjs`. It writes a `::warning`
annotation and a step-summary table row for each test that passed only
on a retry.
- `scripts/run-vitest-stable.mjs` passes `--retry` to `vitest run`. The
`vitest list` collection step does not retry.
- The root `vitest.config.ts` and
`packages/paperclip-runner/vitest.config.ts` add the retry reporter
beside the existing reporters. The Runner config sets `retry` from the
same gate.
- `.github/workflows/pr-trusted.yml` reads the `CI_VITEST_RETRY`
repository variable. Set it to `0` to turn retries off without a code
change.
- Add `scripts/ci/retry-reporter.test.mjs` to `test:release-registry`,
and document the behavior in `doc/DEVELOPING.md`.

## Verification

- `node --test scripts/ci/retry-reporter.test.mjs`: 11/11 pass. The
tests cover the gate matrix, the reporter output, a real Vitest run with
a test that fails once, the wrapper arguments, and the Runner config.
- `node --test scripts/__tests__/run-vitest-stable-shard.test.mjs`:
39/39 pass.
- `node --test .github/scripts/tests/*.test.mjs`: 744/744 pass.
- Manual check: a temporary test that fails once ran through
`run-vitest-stable.mjs` and through the Runner package config. With the
PR environment it passed and wrote the annotation and the summary row.
With `GITHUB_WORKFLOW=Release` it failed. The temporary test is removed.

## Risks

- A retry can hide a flaky test. The annotation and the summary row
reduce this risk.
- A test that fails for a real reason now runs twice in PR CI, so a
failed run takes slightly longer.
- `pr.yml` loads `pr-trusted.yml` from master, so the `CI_VITEST_RETRY`
switch works only after merge. The retry itself is in the scripts and
starts with this pull request.
- Low risk otherwise: release and master verification are not changed.

## Model Used

- Claude Opus 5.5 (`claude-opus-5-5`) in Claude Code, with tool use and
code execution. A Claude sub-agent wrote the change, and the parent
session reviewed it.

## 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

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

Co-authored-by: Claude Opus 5.5 <[email protected]>
2026-11-06 16:43:02 -08:00

195 lines
11 KiB
JavaScript

import assert from "node:assert/strict";
import { spawnSync } from "node:child_process";
import { chmodSync, mkdirSync, mkdtempSync, readFileSync, realpathSync, rmSync, symlinkSync, writeFileSync } from "node:fs";
import { tmpdir } from "node:os";
import path from "node:path";
import { fileURLToPath } from "node:url";
import test from "node:test";
import RetryReporter, { annotation, flakyRecord, summaryMarkdown } from "./retry-reporter.mjs";
import { retryReporterEnabled, vitestRetryArgs, vitestRetryCount } from "./vitest-retry.mjs";
const root = fileURLToPath(new URL("../../", import.meta.url));
const reporterPath = fileURLToPath(new URL("./retry-reporter.mjs", import.meta.url));
const pr = { GITHUB_ACTIONS: "true", GITHUB_WORKFLOW: "PR" };
function testCase({ state = "passed", retryCount = 0, file = path.join(root, "server/src/a.test.ts"), name = "suite > case", line } = {}) {
return {
fullName: name,
module: { moduleId: file },
project: { name: "@paperclipai/server" },
location: line === undefined ? undefined : { line, column: 3 },
result: () => ({ state }),
diagnostic: () => ({ retryCount }),
};
}
test("retry defaults to one only in the PR workflow on GitHub Actions", () => {
assert.equal(vitestRetryCount({}), 0);
assert.equal(vitestRetryCount(pr), 1);
for (const workflow of ["Release", "Release Verify", "Cloud readiness", "Release Smoke", "Trusted PR CI", "Incremental CI qualification"]) {
assert.equal(vitestRetryCount({ GITHUB_ACTIONS: "true", GITHUB_WORKFLOW: workflow }), 0, workflow);
}
assert.equal(vitestRetryCount({ GITHUB_WORKFLOW: "PR" }), 0, "a local shell that happens to set the name stays at 0");
assert.deepEqual(vitestRetryArgs({}), []);
assert.deepEqual(vitestRetryArgs(pr), ["--retry=1"]);
});
test("PAPERCLIP_VITEST_RETRY overrides the default and rejects invalid values", () => {
assert.equal(vitestRetryCount({ ...pr, PAPERCLIP_VITEST_RETRY: "0" }), 0);
assert.deepEqual(vitestRetryArgs({ ...pr, PAPERCLIP_VITEST_RETRY: "0" }), []);
assert.equal(vitestRetryCount({ ...pr, PAPERCLIP_VITEST_RETRY: "" }), 1, "an unset repository variable keeps the PR default");
assert.equal(vitestRetryCount({ ...pr, PAPERCLIP_VITEST_RETRY: " " }), 1);
assert.equal(vitestRetryCount({ PAPERCLIP_VITEST_RETRY: "2" }), 2);
for (const value of ["-1", "1.5", "yes"]) {
assert.throws(() => vitestRetryCount({ PAPERCLIP_VITEST_RETRY: value }), /non-negative integer/, value);
}
});
test("the reporter loads in GitHub Actions jobs and wherever retry is on", () => {
assert.equal(retryReporterEnabled({}), false);
assert.equal(retryReporterEnabled({ GITHUB_ACTIONS: "true", GITHUB_WORKFLOW: "Release" }), true);
assert.equal(retryReporterEnabled({ PAPERCLIP_VITEST_RETRY: "1" }), true);
});
test("only passes that needed a retry are flaky records", () => {
assert.equal(flakyRecord(testCase()), null);
assert.equal(flakyRecord(testCase({ state: "failed", retryCount: 1 })), null);
assert.equal(flakyRecord(testCase({ state: "skipped", retryCount: 1 })), null);
assert.deepEqual(flakyRecord(testCase({ retryCount: 1, line: 12 })), {
file: "server/src/a.test.ts", line: 12, name: "suite > case", project: "@paperclipai/server", retries: 1,
});
});
test("annotations escape workflow-command separators", () => {
const record = flakyRecord(testCase({ retryCount: 2, file: path.join(root, "ui/src/a,b:c.test.ts"), name: "50% > line\nbreak" }));
assert.equal(record.line, null);
assert.equal(annotation(record),
"::warning file=ui/src/a%2Cb%3Ac.test.ts,title=Flaky test passed on retry::50%25 > line%0Abreak (retried 2)");
assert.equal(annotation({ ...record, line: 7 }),
"::warning file=ui/src/a%2Cb%3Ac.test.ts,line=7,title=Flaky test passed on retry::50%25 > line%0Abreak (retried 2)");
});
test("the step summary is a Markdown table with the job name", () => {
assert.equal(summaryMarkdown([]), "");
const markdown = summaryMarkdown([{ file: "server/src/a.test.ts", name: "a | b\nc", retries: 1 }], "General tests (server 1/4)");
assert.equal(markdown, [
"",
"### Flaky tests passed on retry",
"",
"| File | Test | Retries | Job |",
"| --- | --- | --- | --- |",
"| `server/src/a.test.ts` | a \\| b c | 1 | General tests (server 1/4) |",
"", ""].join("\n"));
assert.match(summaryMarkdown([{ file: "f", name: "n", retries: 1 }]), /\| 1 \| - \|$/m);
});
test("the reporter writes nothing for a clean run and both outputs for a flaky one", () => {
const directory = mkdtempSync(path.join(tmpdir(), "retry-reporter-"));
try {
const summary = path.join(directory, "summary.md");
let output = "";
const reporter = new RetryReporter({ env: { GITHUB_STEP_SUMMARY: summary, GITHUB_JOB: "general" }, root, write: (text) => { output += text; } });
reporter.onTestRunStart();
reporter.onTestCaseResult(testCase());
reporter.onTestRunEnd();
assert.equal(output, "");
assert.throws(() => readFileSync(summary), { code: "ENOENT" });
reporter.onTestRunStart();
reporter.onTestCaseResult(testCase({ retryCount: 1 }));
reporter.onTestCaseResult(testCase({ state: "failed", retryCount: 1, name: "still failing" }));
reporter.onTestRunEnd();
assert.equal(output, "::warning file=server/src/a.test.ts,title=Flaky test passed on retry::suite > case (retried 1)\n");
assert.match(readFileSync(summary, "utf8"), /^\| `server\/src\/a\.test\.ts` \| suite > case \| 1 \| general \|$/m);
} finally { rmSync(directory, { recursive: true, force: true }); }
});
function runFlakyFixture(env) {
const directory = realpathSync(mkdtempSync(path.join(tmpdir(), "retry-reporter-vitest-")));
symlinkSync(path.join(root, "node_modules"), path.join(directory, "node_modules"), "dir");
// The counter lives on disk because a retry reruns the case in the same worker.
writeFileSync(path.join(directory, "flaky.test.mjs"), `import { existsSync, writeFileSync } from "node:fs";
import { test } from "vitest";
test("fails on its first attempt", () => {
const marker = new URL("./attempted", import.meta.url);
if (!existsSync(marker)) { writeFileSync(marker, ""); throw new Error("first attempt fails"); }
});
test("always passes", () => {});
`);
const configFile = path.join(directory, "vitest.config.mjs");
writeFileSync(configFile, `export default ${JSON.stringify({ test: { include: ["*.test.mjs"], reporters: ["default", reporterPath] } })};`);
const summary = path.join(directory, "summary.md");
const childEnv = { ...process.env, ...env, GITHUB_STEP_SUMMARY: summary, GITHUB_WORKSPACE: directory, PAPERCLIP_CI_PROFILE_JOB: "probe" };
delete childEnv.NODE_TEST_CONTEXT;
const result = spawnSync(process.execPath, [path.join(root, "node_modules/vitest/vitest.mjs"), "run",
"--config", configFile, "--root", directory, ...vitestRetryArgs(childEnv)], {
cwd: directory, encoding: "utf8", timeout: 60_000, env: childEnv,
});
let summaryText = "";
try { summaryText = readFileSync(summary, "utf8"); } catch {}
rmSync(directory, { recursive: true, force: true });
return { result, summaryText };
}
test("a real PR-gated Vitest run retries the flake once and reports it", () => {
const { result, summaryText } = runFlakyFixture({ ...pr, PAPERCLIP_VITEST_RETRY: "" });
assert.equal(result.status, 0, result.stdout + result.stderr);
assert.match(result.stdout, /^::warning file=flaky\.test\.mjs,title=Flaky test passed on retry::fails on its first attempt \(retried 1\)$/m);
assert.doesNotMatch(result.stdout, /always passes \(retried/);
assert.match(summaryText, /^\| `flaky\.test\.mjs` \| fails on its first attempt \| 1 \| probe \|$/m);
});
test("outside the PR workflow the same flake fails without a retry", () => {
const { result, summaryText } = runFlakyFixture({ GITHUB_ACTIONS: "true", GITHUB_WORKFLOW: "Release", PAPERCLIP_VITEST_RETRY: "" });
assert.equal(result.status, 1, result.stdout + result.stderr);
assert.doesNotMatch(result.stdout, /Flaky test passed on retry/);
assert.doesNotMatch(summaryText, /Flaky tests passed on retry/);
});
test("run-vitest-stable passes --retry to vitest run only in the PR workflow", () => {
const directory = mkdtempSync(path.join(tmpdir(), "retry-wrapper-"));
try {
const bin = path.join(directory, "bin");
mkdirSync(bin);
writeFileSync(path.join(bin, "pnpm"), '#!/bin/sh\nprintf "%s\\n" "$@" >> "$RETRY_ARGS_FILE"\n');
chmodSync(path.join(bin, "pnpm"), 0o755);
const invoke = (env) => {
const argsFile = path.join(directory, `args-${Math.random()}`);
const childEnv = { ...process.env, ...env, PATH: `${bin}${path.delimiter}${process.env.PATH}`, RETRY_ARGS_FILE: argsFile };
for (const key of Object.keys(childEnv)) {
if (key === "PAPERCLIP_CI_TEST_SELECTION" || key === "PAPERCLIP_CI_PROFILE" || /^PAPERCLIP_.*_TEST_DATABASE_URL$/.test(key)) delete childEnv[key];
}
const result = spawnSync(process.execPath, [path.join(root, "scripts/run-vitest-stable.mjs"),
"--mode", "general", "--group", "general-workspaces-c"], { cwd: root, encoding: "utf8", env: childEnv, timeout: 60_000 });
assert.equal(result.status, 0, result.stdout + result.stderr);
return readFileSync(argsFile, "utf8").trim().split("\n");
};
assert.deepEqual(invoke({ ...pr, PAPERCLIP_VITEST_RETRY: "" }),
["exec", "vitest", "run", "--exclude", "**/dist/**", "--retry=1", "--project", "@paperclipai/db"]);
assert.deepEqual(invoke({ ...pr, PAPERCLIP_VITEST_RETRY: "0" }),
["exec", "vitest", "run", "--exclude", "**/dist/**", "--project", "@paperclipai/db"]);
assert.deepEqual(invoke({ GITHUB_ACTIONS: "true", GITHUB_WORKFLOW: "Cloud readiness", PAPERCLIP_VITEST_RETRY: "" }),
["exec", "vitest", "run", "--exclude", "**/dist/**", "--project", "@paperclipai/db"]);
} finally { rmSync(directory, { recursive: true, force: true }); }
});
test("the Runner package config retries only in the PR workflow", () => {
const base = { ...process.env };
for (const key of ["GITHUB_ACTIONS", "GITHUB_WORKFLOW", "PAPERCLIP_VITEST_RETRY", "PAPERCLIP_CI_PROFILE"]) delete base[key];
const load = (env) => {
const childEnv = { ...base, ...env };
const result = spawnSync(process.execPath, ["--input-type=module", "-e",
"const { default: config } = await import('./packages/paperclip-runner/vitest.config.ts');"
+ "process.stdout.write(JSON.stringify({ retry: config.test.retry ?? 0, reporters: config.test.reporters ?? [] }));"],
{ cwd: root, encoding: "utf8", env: childEnv, timeout: 60_000 });
assert.equal(result.status, 0, result.stderr);
return JSON.parse(result.stdout);
};
const prConfig = load({ ...pr, PAPERCLIP_VITEST_RETRY: "" });
assert.equal(prConfig.retry, 1);
assert.ok(prConfig.reporters.includes(reporterPath), JSON.stringify(prConfig.reporters));
assert.equal(load({ GITHUB_ACTIONS: "true", GITHUB_WORKFLOW: "Release", PAPERCLIP_VITEST_RETRY: "" }).retry, 0);
assert.equal(load({ ...pr, PAPERCLIP_VITEST_RETRY: "0" }).retry, 0);
assert.deepEqual(load({}), { retry: 0, reporters: [] }, "local runs keep the unmodified config");
});