Files
PaperClipAI/.github/scripts/tests/ci-profile.test.mjs
T
DottaandPaperclip 2cc7992369 Share the external Runner lifecycle and add strict Muse contracts (#16444)
Add the external Runner foundation with distinct Dot/Grok V6, canonical home-mode V7 and Muse V8 contracts.

Keep generated schemas and profiles synchronized. Bound default question-answer command IDs in a disjoint namespace while preserving existing IDs and durable replay. Preserve canonical migration history and add normally generated, replay-safe Muse schema milestones with literal preview-history fixtures.

Refs #16444.

Co-Authored-By: Paperclip <noreply@paperclip.ing>
2026-10-16 13:46:50 -05:00

356 lines
20 KiB
JavaScript

import assert from "node:assert/strict";
import test from "node:test";
import { mkdtempSync, mkdirSync, readFileSync, existsSync, rmSync, symlinkSync, writeFileSync } from "node:fs";
import { execFileSync, spawnSync } from "node:child_process";
import { tmpdir } from "node:os";
import path from "node:path";
import { fileURLToPath } from "node:url";
import { buildTrace } from "../../../scripts/ci-profile/trace.mjs";
import { sendEvents } from "../../../scripts/ci-profile/export.mjs";
import { associatedProfilePr, ciOwners, trustedOwnerIds } from "../../../scripts/ci-profile/authorize.mjs";
const at = (seconds) => new Date(1_800_000_000_000 + seconds * 1000).toISOString();
const run = {
id: 42, run_attempt: 1, run_number: 7, status: "completed", conclusion: "failure",
repository: { id: 1170821064, full_name: "paperclipai/paperclip" }, head_repository: { id: 99 },
actor: { id: 123, type: "User" }, triggering_actor: { id: 123, type: "User" },
head_sha: "abc", head_branch: "branch", created_at: at(0), run_started_at: at(0),
html_url: "https://github.com/paperclipai/paperclip/actions/runs/42",
};
const job = (id, name, start, end) => ({
id, run_id: 42, run_attempt: 1, name: `ci / ${name}`, status: "completed",
conclusion: "success", created_at: at(1), started_at: at(start), completed_at: at(end),
labels: ["ubuntu-latest"], steps: [
{ number: 1, name: "Initialize CI profile", started_at: at(start), completed_at: at(start + 1), conclusion: "success" },
{ number: 2, name: "Tests", started_at: at(start + 1), completed_at: at(end), conclusion: "success" },
],
});
const jobs = [job(1, "first", 2, 12), job(2, "second", 3, 20)];
const record = (fields) => ({ version: 1, runId: "42", attempt: "1", invocation: "invocation", job: "first", framework: "vitest", ...fields });
test("one wall-clock trace includes parallel jobs, wait, and every step", () => {
const trace = buildTrace(run, jobs);
assert.equal(trace.wallMs, 20000);
assert.equal(trace.events[0].data.duration_ms, 20000);
assert.equal(trace.events.filter((event) => event.data["ci.kind"] === "job").length, 2);
assert.equal(trace.events.filter((event) => event.data["ci.kind"] === "step").length, 4);
assert.equal(trace.events.filter((event) => event.data["ci.kind"] === "wait").length, 2);
assert.equal(new Set(trace.events.map((event) => event.data["trace.trace_id"])).size, 1);
assert.equal(trace.events.find((event) => event.data.name === "ci / second").data["ci.finishes_last"], true);
assert.equal(trace.events[0].data.error, true);
});
test("file/test hierarchy keeps runner timestamps and rejects unrelated artifact data", () => {
const records = [
record({ kind: "job" }), record({ kind: "job", job: "second" }),
record({ kind: "start" }), record({ kind: "end" }),
record({ kind: "file", id: "file", name: "a.test.ts", file: "a.test.ts", start: Date.parse(at(4)), duration: 6000, status: "passed" }),
record({ kind: "test", id: "case", parent: "file", name: "runs", file: "a.test.ts", start: Date.parse(at(5)), duration: 100, status: "passed", secret: "must not leave artifact" }),
record({ kind: "test", id: "foreign", runId: "41", name: "wrong run" }),
];
const trace = buildTrace(run, jobs, records);
const file = trace.events.find((event) => event.data.name === "a.test.ts");
const entry = trace.events.find((event) => event.data.name === "runs");
assert.equal(entry.time, at(5));
assert.equal(entry.data["trace.parent_id"], file.data["trace.span_id"]);
assert.equal(file.data["trace.parent_id"], trace.events.find((event) => event.data.name === "Tests").data["trace.span_id"]);
assert.equal(trace.events.some((event) => event.data.name === "wrong run"), false);
assert.doesNotMatch(JSON.stringify(trace), /must not leave artifact/);
assert.equal(trace.events[0].data["ci.profile_complete"], false);
assert.match(trace.warnings.join(), /1 invalid/);
});
test("short files belong to the latest containing step despite second-precision boundaries", () => {
const records = [record({ kind: "file", id: "file", name: "short.test.ts", file: "short.test.ts",
start: Date.parse(at(3)) + 100, duration: 50, status: "passed" })];
const trace = buildTrace(run, jobs, records);
const file = trace.events.find((event) => event.data.name === "short.test.ts");
const step = trace.events.find((event) => event.data.name === "Tests");
assert.equal(file.data["trace.parent_id"], step.data["trace.span_id"]);
});
test("cancelled reporters and missing shards remain explicit", () => {
const trace = buildTrace({ ...run, conclusion: "cancelled" }, jobs, [record({ kind: "job" }), record({ kind: "start" })]);
assert.equal(trace.events[0].data["ci.profile_complete"], false);
assert.match(trace.warnings.join(), /Missing timing artifacts for 1 jobs/);
assert.match(trace.warnings.join(), /1 test invocations/);
});
test("a completely absent reporter invocation cannot hide behind a successful sibling", () => {
const records = [
record({ kind: "job" }), record({ kind: "job", job: "second" }),
record({ kind: "expected" }), record({ kind: "start" }), record({ kind: "end" }),
record({ kind: "file", id: "file", name: "a.test.ts", file: "a.test.ts", start: Date.parse(at(4)), duration: 100, status: "passed" }),
record({ kind: "expected", invocation: "reporter-never-loaded" }),
];
const trace = buildTrace(run, jobs, records);
assert.equal(trace.events[0].data["ci.profile_complete"], false);
assert.equal(trace.events[0].data["ci.expected_invocations"], 2);
assert.equal(trace.events[0].data["ci.reported_invocations"], 1);
assert.match(trace.warnings.join(), /1 test invocations/);
});
test("coverage is conservative after failures and does not merge framework invocation IDs", () => {
const records = [
record({ kind: "job" }), record({ kind: "job", job: "second" }),
record({ kind: "expected" }), record({ kind: "start" }), record({ kind: "end" }),
];
assert.equal(buildTrace({ ...run, conclusion: "success" }, jobs, records).events[0].data["ci.profile_complete"], true);
assert.equal(buildTrace(run, jobs, records).events[0].data["ci.profile_complete"], false);
const trace = buildTrace({ ...run, conclusion: "success" }, jobs, [
...records, record({ framework: "node", kind: "expected" }),
]);
assert.equal(trace.events[0].data["ci.profile_complete"], false);
assert.equal(trace.events[0].data["ci.expected_invocations"], 2);
assert.equal(trace.events[0].data["ci.reported_invocations"], 1);
});
test("complete reporter pairs count without launcher intent, and long test names remain timed", () => {
const records = [
record({ kind: "job" }), record({ kind: "job", job: "second" }),
record({ framework: "node", kind: "start" }), record({ framework: "node", kind: "end" }),
record({ kind: "test", id: "long", parent: "file", name: "x".repeat(5000),
file: "a.test.ts", start: Date.parse(at(4)), duration: 100, status: "passed" }),
record({ kind: "file", id: "file", name: "a.test.ts", file: "a.test.ts",
start: Date.parse(at(4)), duration: 100, status: "passed" }),
];
const trace = buildTrace({ ...run, conclusion: "success" }, jobs, records);
assert.equal(trace.events[0].data["ci.profile_complete"], true);
assert.equal(trace.events[0].data["ci.test_count"], 1);
assert.equal(trace.events.find((event) => event.data["ci.kind"] === "test").data.name.length, 256);
});
test("trusted policy commands work with both current and pre-profile checkouts", () => {
const source = readFileSync(new URL("../../workflows/pr-trusted.yml", import.meta.url), "utf8");
const block = source.split(" - name: Select compatible Node test command\n")[1].split("\n - name:")[0];
const script = block.split(" run: |\n")[1].split("\n").map((line) => line.replace(/^ {10}/, "")).join("\n");
const directory = mkdtempSync(path.join(tmpdir(), "ci-profile-legacy-"));
try {
writeFileSync(path.join(directory, "probe.test.cjs"), 'require("node:test")("old checkout works", () => {});');
for (const current of [false, true]) {
if (current) {
mkdirSync(path.join(directory, "scripts"));
symlinkSync(fileURLToPath(new URL("../../../scripts/ci-profile", import.meta.url)), path.join(directory, "scripts/ci-profile"));
}
const env = { ...process.env, GITHUB_ENV: path.join(directory, "env"), PAPERCLIP_CI_PROFILE: "false" };
delete env.NODE_TEST_CONTEXT;
const result = spawnSync("bash", ["-c", `${script}\nsource "$GITHUB_ENV"\nnode "$CI_NODE_TEST" probe.test.cjs`], {
cwd: directory, env, encoding: "utf8",
});
assert.equal(result.status, 0, result.stdout + result.stderr);
assert.match(readFileSync(env.GITHUB_ENV, "utf8"), current ? /CI_NODE_TEST=scripts\/ci-profile\/node-test.mjs/ : /CI_NODE_TEST=--test/);
}
} finally { rmSync(directory, { recursive: true, force: true }); }
});
test("reruns use separate IDs and clocks and cannot mix job attempts", () => {
const rerun = { ...run, run_attempt: 2, run_started_at: at(2) };
assert.throws(() => buildTrace(rerun, jobs), /another run or attempt/);
const trace = buildTrace(rerun, jobs.map((entry) => ({ ...entry, run_attempt: 2 })));
assert.equal(trace.wallMs, 18000);
assert.notEqual(trace.traceId, buildTrace(run, jobs).traceId);
assert.equal(trace.events[0].data["ci.partial_rerun"], true);
assert.throws(() => buildTrace({ ...run, status: "in_progress" }, jobs), /complete/);
});
test("partial reruns exclude old jobs that GitHub relabels with the new attempt number", () => {
const rerun = { ...run, run_attempt: 2, run_started_at: at(30) };
const inventory = [
{ ...job(1, "carried build", 2, 12), run_attempt: 2 },
{ ...job(3, "retried tests", 32, 50), run_attempt: 2, created_at: at(31) },
];
const trace = buildTrace(rerun, inventory);
assert.equal(trace.wallMs, 20000);
assert.equal(trace.jobCount, 1);
assert.equal(trace.events[0].data["ci.carried_job_count"], 1);
assert.equal(trace.events.some((event) => event.data.name === "ci / carried build"), false);
assert.equal(trace.events[0].data["ci.profile_complete"], false);
assert.match(trace.warnings.join(), /1 completed jobs carried over/);
assert.throws(() => buildTrace(rerun, inventory.slice(0, 1)), /No jobs executed/);
});
test("label alone cannot authorize outsiders, bots, stale revisions or outsider reruns", () => {
const pr = { user: { id: 123, type: "User" }, base: { repo: run.repository }, head: { repo: { id: 99 }, sha: "abc" }, labels: [{ name: "profile-ci" }] };
assert.equal(associatedProfilePr(run, [pr], [123]), pr);
for (const changed of [
{ ...pr, labels: [] },
{ ...pr, head: { ...pr.head, sha: "next" } },
{ ...pr, head: { ...pr.head, repo: { id: 100 } } },
{ ...pr, user: { id: 999, type: "User" } },
{ ...pr, user: { id: 123, type: "Bot" } },
{ ...pr, base: { repo: { ...run.repository, id: 100 } } },
]) assert.equal(associatedProfilePr(run, [changed], [123]), undefined);
assert.equal(associatedProfilePr({ ...run, actor: { id: 999, type: "User" } }, [pr], [123]), undefined);
assert.equal(associatedProfilePr({ ...run, triggering_actor: { id: 999, type: "User" } }, [pr], [123]), undefined);
assert.equal(associatedProfilePr({ ...run, repository: { ...run.repository, id: 100 } }, [pr], [123]), undefined);
assert.equal(associatedProfilePr(run, [pr], []), undefined);
});
test("profiling uses only infrastructure CODEOWNERS resolved at trusted master", () => {
const source = '.github/** @owner @second\ndoc/** @outsider\n';
assert.deepEqual(ciOwners(source), ["owner", "second"]);
for (const invalid of ['', '.github/**', '.github/** @org/team', '.github/** @owner\n/.github/** @other']) {
assert.throws(() => ciOwners(invalid), /CODEOWNERS/);
}
const endpoints = [];
const sha = "a".repeat(40);
assert.deepEqual(trustedOwnerIds((endpoint) => {
endpoints.push(endpoint);
if (endpoint.endsWith('git/ref/heads/master')) return { object: { sha } };
if (endpoint.endsWith(`CODEOWNERS?ref=${sha}`)) return { encoding: 'base64', content: Buffer.from(source).toString('base64') };
return { id: endpoint === '/users/owner' ? 123 : 456, type: 'User' };
}), [123, 456]);
assert.deepEqual(endpoints, [
'/repos/paperclipai/paperclip/git/ref/heads/master',
`/repos/paperclipai/paperclip/contents/.github/CODEOWNERS?ref=${sha}`, '/users/owner', '/users/second',
]);
});
test("export paces batches and retries only rejected transient events", async () => {
const events = Array.from({ length: 501 }, () => ({ data: {} }));
const sizes = [];
const pauses = [];
await sendEvents(events, "test-only", "paperclip-ci", async (url, options) => {
assert.equal(url, "https://api.honeycomb.io/1/batch/paperclip-ci");
assert.equal(options.redirect, "error");
const size = JSON.parse(options.body).length;
sizes.push(size);
return { ok: true, json: async () => Array.from({ length: size }, () => ({ status: 202 })) };
}, async (ms) => { pauses.push(ms); });
assert.deepEqual(sizes, [250, 250, 1]);
assert.deepEqual(pauses, [1000, 1000]);
let calls = 0;
const retried = [];
await sendEvents([{ data: { id: 1 } }, { data: { id: 2 } }], "test-only", "paperclip-ci", async (_url, options) => {
const batch = JSON.parse(options.body);
retried.push(batch.map((event) => event.data.id));
calls += 1;
return { ok: true, json: async () => calls === 1 ? [{ status: 202 }, { status: 429 }] : [{ status: 202 }] };
}, async (ms) => { pauses.push(ms); });
assert.deepEqual(retried, [[1, 2], [2]]);
assert.equal(pauses.at(-1), 1000);
await assert.rejects(sendEvents([events[0]], "test-only", "paperclip-ci", async () => ({ ok: true, json: async () => [{ status: 400 }] })), /status 400/);
});
function secretReference(text, includeQuoted = false) {
for (const line of text.split(/[\r\n\u2028\u2029]/)) {
const start = line.indexOf("HONEYCOMB");
if (start >= 0 && line.indexOf("KEY", start + "HONEYCOMB".length) >= 0) return true;
}
let offset = 0;
while (offset < text.length) {
const start = text.indexOf("${{", offset);
if (start < 0) return false;
offset = start + 3;
let quoted = false;
let expression = "";
while (offset < text.length) {
const char = text[offset];
if (char === "'") {
if (quoted && text[offset + 1] === "'") {
offset += 2;
continue;
}
quoted = !quoted;
expression += includeQuoted ? char : " ";
} else if (!quoted) {
if (char === "}" && text[offset + 1] === "}") {
offset += 2;
break;
}
expression += char;
} else if (includeQuoted) {
expression += char;
}
offset += 1;
}
if (/\bsecrets\b/.test(expression)) return true;
}
return false;
}
function assertNoPrWorkflowCredentials(source) {
assert.equal(secretReference(source, true), false);
}
test("PR credential guard permits filenames but rejects secret expressions", () => {
assertNoPrWorkflowCredentials("run: node --test scripts/docker-compose-secrets.test.mjs");
assertNoPrWorkflowCredentials("if: ${{ always() }}\nrun: node --test scripts/docker-compose-secrets.test.mjs");
assertNoPrWorkflowCredentials("if: ${{ startsWith(github.ref, 'ref') }}\nrun: 'node --test scripts/docker-compose-secrets.test.mjs'");
for (const source of ["env: ${{ secrets.EXAMPLE }}", "env: ${{ secrets['EXAMPLE'] }}", "env: ${{ toJSON(secrets) }}", "env: ${{ format('}}', secrets.EXAMPLE) }}",
"# Expressions start with '${{'.\nTOKEN: ${{ secrets.PR_TOKEN }}",
"name: Example # Expressions start with '${{'.\nTOKEN: ${{ secrets.PR_TOKEN }}"]) {
assert.throws(() => assertNoPrWorkflowCredentials(source));
}
for (const source of [
"${{secrets.EXAMPLE_TOKEN}}", '${{ secrets["EXAMPLE_TOKEN"] }}',
"${{ failure() && secrets.EXAMPLE_TOKEN }}", "${{\n secrets.EXAMPLE_TOKEN\n}}",
"${{ format('{0}', secrets.EXAMPLE_TOKEN) }}", "HONEYCOMB_CI_API_KEY",
]) assert.throws(() => assertNoPrWorkflowCredentials(source));
assert.throws(() => assertNoPrWorkflowCredentials("HONEYCOMB_API_KEY: example"));
assert.throws(() => assertNoPrWorkflowCredentials("${{ format('secrets.EXAMPLE', 'literal') }}"));
});
test("workflow exporter runs trusted code and keeps keys out of the PR jobs", () => {
const workflow = readFileSync(new URL("../../workflows/ci-profile.yml", import.meta.url), "utf8");
assert.match(workflow, /ref: \$\{\{ github.sha \}\}/);
assert.match(workflow, /startsWith\(github.event.workflow_run.display_title/);
assert.match(workflow, /environment: ci-profiling/);
assert.doesNotMatch(workflow, /head_sha|secrets: inherit|contents: write|pull-requests: write|pnpm install/);
const pr = readFileSync(new URL("../../workflows/pr-trusted.yml", import.meta.url), "utf8");
assert.equal(secretReference("run: node --test scripts/docker-compose-secrets.test.mjs"), false);
for (const lookup of ["${{ secrets.PR_TOKEN }}", "${{ secrets['PR_TOKEN'] }}", "${{\n secrets.PR_TOKEN\n }}", "${{ toJSON(secrets) }}", "${{ secrets.HONEYCOMB_API_KEY }}"]) {
assert.equal(secretReference(lookup), true);
}
for (const lookup of [
"${{ format('Bearer {0}', secrets.PR_TOKEN) }}",
"${{ format('{0}', toJSON(secrets)) }}",
"${{ format('{{Bearer}} {0}', secrets['PR_TOKEN']) }}",
"${{ format('it''s {0}',\n secrets.PR_TOKEN) }}",
]) {
assert.equal(secretReference(lookup), true);
}
for (const nonSecret of [
"${{ format('secrets.PR_TOKEN', 'literal') }}",
"${{ github.sha }}\nrun: node --test scripts/docker-compose-secrets.test.mjs",
]) {
assert.equal(secretReference(nonSecret), false);
}
const escapedExpressions = [28, 10_000].flatMap(count => {
const expression = "${{ format('" + "it''s ".repeat(count) + "{0}', github.sha) }}";
return [expression, expression.replace("github.sha", "secrets.PR_TOKEN"), expression.replace("github.sha", "toJSON(secrets)")];
});
const boundedScan = spawnSync(process.execPath, ["-e", `const fs=require('node:fs');const scan=${secretReference.toString()};process.stdout.write(JSON.stringify(JSON.parse(fs.readFileSync(0,'utf8')).map(value=>[scan(value),scan(value,true)])));`], {
input: JSON.stringify(escapedExpressions),
encoding: "utf8", timeout: 2_000,
});
assert.equal(boundedScan.status, 0, boundedScan.error?.message);
assert.deepEqual(JSON.parse(boundedScan.stdout), [[false, false], [true, true], [true, true], [false, false], [true, true], [true, true]]);
assert.equal(secretReference(pr), false);
assertNoPrWorkflowCredentials(pr);
assert.match(pr, /if: always\(\) && env.PAPERCLIP_CI_PROFILE == 'true'/);
assert.doesNotMatch(pr, /PAPERCLIP_CI_PROFILE:.*labels/);
assert.equal(pr.match(/PAPERCLIP_CI_PROFILE: \$\{\{ needs.gate.outputs.profile_ci \}\}/g).length, 9);
// upload-artifact excludes files inside dot-directories by default. Keep
// this opt-in scoped to timing NDJSON, never the workspace or other secrets.
assert.equal(pr.match(/path: \.ci-profile\/\*\.ndjson\n {10}include-hidden-files: true/g).length, 9);
const routing = readFileSync(new URL("../../workflows/ci-routing.yml", import.meta.url), "utf8");
assert.match(routing, /PROFILE_CI: \$\{\{ steps.profile.outputs.profile_ci \}\}/);
assert.match(routing, /\/git\/ref\/heads\/master/);
assert.match(routing, /authorize\.mjs\?ref=\$TRUSTED_PROFILE_SHA/);
});
test("artifact parsing never extracts paths or executes files and rejects oversized records", () => {
const directory = mkdtempSync(path.join(tmpdir(), "ci-artifact-test-"));
const archive = path.join(directory, "artifact.zip");
const reader = fileURLToPath(new URL("../../../scripts/ci-profile/read-artifact.py", import.meta.url));
try {
execFileSync("python3", ["-c", "import zipfile,sys; z=zipfile.ZipFile(sys.argv[1],'w'); z.writestr('../escape.ndjson', '{\"kind\":\"test\"}\\n'); z.close()", archive]);
const result = execFileSync("python3", [reader, archive], { encoding: "utf8" });
assert.deepEqual(JSON.parse(result), { kind: "test" });
assert.equal(existsSync(path.join(directory, "escape.ndjson")), false);
execFileSync("python3", ["-c", "import zipfile,sys; z=zipfile.ZipFile(sys.argv[1],'w'); z.writestr('large.ndjson', 'x'*33000); z.close()", archive]);
assert.notEqual(spawnSync("python3", [reader, archive]).status, 0);
} finally { rmSync(directory, { recursive: true, force: true }); }
});