Files
msd-core/tests/graphify-auto-update.slow.test.cjs
Tom Boucher 5fd5c81042 test(#3055): add the process seam so a subprocess timeout is expressible as data (#3066)
* test(#3055): add the process seam and route runGsdTools through it

Adds tests/helpers/process-seam.cjs — runNode/runGit/runHook over spawnSync,
each returning a typed discriminated union
{ outcome, exitCode, stdout, stderr, timedOut, signal, killed, code }.
Every call is timeout-bounded; there is no unbounded path.

runGsdTools becomes an adapter over the seam. Its legacy
{ success, output, error, exitCode } shape and retry-once-on-kill behaviour
are preserved byte-identically, so none of its 136 caller files change.

Outcome discrimination was corrected against probed runtime behaviour rather
than assumption: a timeout and a maxBuffer overflow are identical on both
status (null) and signal (SIGTERM), and differ only by code (ETIMEDOUT vs
ENOBUFS). Overflow is therefore classified before timeout. This fixes a live
defect — the previous isKilled() treated an overflow as a kill, retried it for
a second full 60s run, and then reported "host OOM or scheduler contention"
for a child that had merely printed too much.

Also widens the ESLint tests glob from tests/**/*.test.cjs to tests/**/*.cjs,
which brought 31 previously unlinted shared helpers under the same rules their
sibling test files already obey, and fixes the 5 violations that surfaced —
including a bare npm invocation without shell:true in
tests/helpers/emitted-runtime.cjs (DEFECT.WINDOWS-TEST-PORTABILITY), now
routed through the existing portable runNpm helper.

Refs #3051

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* test(#3055): migrate every local spawn wrapper onto the process seam

Replaces the spawn body of all 25 local runHook/runGuard/runGate definitions
with a call to tests/helpers/process-seam.cjs. Each wrapper keeps its name,
parameter list, return shape and post-processing (JSON parse, ANSI strip, env
sanitising, field extraction) — only the spawn mechanism changes, so no test
assertion moves.

The 4 bash-driven wrappers use the seam's explicit `interpreter` option rather
than a fourth primitive; it is explicit rather than inferred from the file
extension, because guessing an interpreter from a path fails silently when a
script's name does not match its shebang.

Seven wrappers were previously unbounded and now carry an explicit timeout
sized to what each actually runs, not the seam default. Two of those seven
(gsd-write-guard, lint-docs-command-form) were absent from the issue's
inventory entirely and were found by scanning after the migration.

Adds the CONTEXT.md `### Process seam` glossary entry and a CONTRIBUTING.md
reference section covering the three primitives, the discriminated union, and
the two rules the seam enforces.

Scope disclosure recorded in the phase design notes: the issue scoped three
identifier names. A scan for local helpers that spawn AND return the spawn
result finds 113 across 82 names, 71 of them unbounded, plus 122 unbounded
direct git call sites. This change bounds 25 of those. The remaining surface
is the same defect class and is NOT closed by this PR.

Refs #3051

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(#3055): classify an externally-killed child as KILLED, not EXITED

Blocker found in this branch's own diff, independently confirmed by an
isolated reviewer.

A child killed by an external signal — a genuine bench OOM kill — makes
spawnSync return { status: null, signal: 'SIGKILL' } with NO .error field.
The seam's "no error implies EXITED" rule therefore classified it as a clean
exit, and runGsdTools returned { success: false, exitCode: 1 } without
retrying. That silently defeated the #969 kill-discrimination for precisely
the case it was built for: the old isKilled() fired on `signal != null`,
retried once, then threw a labelled resource-starvation error. A real OOM
would have been reported as an ordinary assertion failure.

Adds a fifth outcome, KILLED, for "no error but a signal is set", and makes
the adapter retry on TIMED_OUT or KILLED — reproducing the old
`killed || signal != null || code === 'ETIMEDOUT'` condition exactly.
SPAWN_FAILED still does not retry (matching the old behaviour, where signal
was null). BUFFER_OVERFLOW still does not retry, which remains a deliberate
divergence: the old code retried it because signal was SIGTERM, burning a
second 60s run on a child that had merely printed too much.

All five outcomes verified against the live runtime rather than assumed:
SIGKILL -> killed, exit 0/7 -> exited, timeout -> timed_out (ETIMEDOUT),
>1MB stdout -> buffer_overflow (ENOBUFS).

Refs #3051

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(#3055): address standards-review findings on this branch

Three findings from the standards axis of the review, all in this branch's
own diff.

The CONTEXT.md glossary entry this branch introduced was already stale on the
branch's own last commit: it enumerated a 4-member OUTCOME while the code had
5, because the KILLED fix did not update it. That is precisely the drift the
"module changes update Domain-terms" gate exists to catch, so the entry now
lists all five and explains KILLED.

api-coverage-gate-e2e compared an outcome against the raw string 'exited'
rather than OUTCOME.EXITED, the only such outlier; the enum is now imported
and used. A sweep for the other four outcome literals found no further
comparison sites.

Three call sites hand the literal bash flag '-c' to the seam's first
parameter, which the JSDoc described as an absolute script path. Rather than
add a fourth primitive, the contract is corrected to match reality: the
parameter is renamed `target` and documented as the first argv element handed
to the interpreter — normally a script path, but for an interpreter invoked
with an inline program it may be that interpreter's own flag. No behaviour
change.

Refs #3051

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* test(#3055): assert the cross-platform timeout contract, not the macOS one

The remote runner failed on both Linux lanes (node 22 and node 24, identical)
while the same tests passed locally on macOS. Two assertions encoded a
platform-specific behaviour as a cross-platform guarantee.

When spawnSync times out, macOS preserves the child's partial stdout/stderr;
Linux discards it and returns empty strings. Verified on node v26.5.1 both
ways. The seam passes through whatever spawnSync hands it and cannot
manufacture output that was discarded, so the production code was correct —
the tests were wrong.

Both tests now assert the guarantee the seam actually makes on every
platform: outcome TIMED_OUT, timedOut true, and stdout/stderr always being
strings rather than undefined or a Buffer. The partial-content assertions are
retained behind an explicit process.platform === 'darwin' guard so the macOS
coverage is not lost, and the first test is renamed to say what it now
guarantees.

This is the failure mode the remote matrix exists to catch: local macOS
verification would have shipped it.

Refs #3051

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(#3055): classify a failed spawn as SPAWN_FAILED, not a timeout

Windows CI caught two defects the Linux matrix could not.

tests/context-predicates-query.test.cjs passes a 32K-char argv value. On
Windows that exceeds the argv limit and spawnSync fails with code
ENAMETOOLONG, signal null, status null. The seam's fallback rule — "otherwise,
status === null implies TIMED_OUT" — swallowed it, so the adapter retried a
spawn that can never succeed and then threw the resource-starvation error. The
old isKilled() returned false for that shape and returned an ordinary failure
result.

TIMED_OUT is now identified positively: code === 'ETIMEDOUT' OR signal is set.
Anything else carrying an error is SPAWN_FAILED, which covers ENAMETOOLONG,
E2BIG, EACCES and ENOENT alike. The signal clause is what keeps a platform
whose timeout errno differs classified correctly, so the greedy catch-all is no
longer needed.

The second defect is a contract regression I introduced and had claimed
otherwise. That same test asserts `typeof r.exitCode === 'number'`, and
toLegacyShape was returning null for BUFFER_OVERFLOW and SPAWN_FAILED, so the
assertion failed on type. The old code returned `err.status ?? 1` on every
non-retried failure path. The adapter now returns 1 again for both, and the
comment claiming "never coerced to exitCode:1, unlike the pre-seam helper" is
retracted: the seam keeps the richer truth (exitCode null plus a distinct
outcome), the legacy adapter keeps the old numeric contract its callers
actually depend on.

Verified on this host: a 4MB argv yields E2BIG -> SPAWN_FAILED; ENOENT ->
SPAWN_FAILED; timeout -> TIMED_OUT; >1MB stdout -> BUFFER_OVERFLOW; SIGKILL ->
KILLED; clean exit -> EXITED.

Refs #3051

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: sim <sim@local>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
2026-08-05 01:06:39 -04:00

789 lines
33 KiB
JavaScript

'use strict';
// Tests for graphify.cjs — auto-update describe block.
// Split from the consolidated 2336-LOC file. Refs #3761.
//
// Regression for #3347: opt-in auto-update of the knowledge graph after
// main HEAD advances. Two sub-concerns: config-key surface (config.test)
// and hook behavior (hook.test). Both merged here.
const { describe, test } = require('node:test');
const assert = require('node:assert/strict');
const fs = require('fs');
const path = require('path');
const os = require('node:os');
const { execFileSync, spawnSync } = require('child_process');
const { createTempProject, cleanup, runGsdTools, delay } = require('./helpers.cjs');
const { runHook: seamRunHook } = require('./helpers/process-seam.cjs');
const {
graphifyStatus,
} = require('../gsd-core/bin/lib/graphify.cjs');
const {
VALID_CONFIG_KEYS,
isValidConfigKey,
} = require('../gsd-core/bin/lib/config-schema.cjs');
const {
CONFIG_DEFAULTS: CANONICAL_CONFIG_DEFAULTS,
} = require('../gsd-core/bin/lib/configuration.cjs');
const {
makeStatusProject,
} = require('./helpers/graphify.cjs');
// ─── auto-update describe ─────────────────────────────────────────────────────
describe('auto-update', () => {
describe('config-key surface', () => {
// Regression for #3347
test('VALID_CONFIG_KEYS contains graphify.auto_update', () => {
assert.ok(
VALID_CONFIG_KEYS.has('graphify.auto_update'),
'graphify.auto_update must be in VALID_CONFIG_KEYS so config-set accepts it',
);
});
test('isValidConfigKey accepts graphify.auto_update', () => {
assert.ok(
isValidConfigKey('graphify.auto_update'),
'isValidConfigKey must return true for graphify.auto_update',
);
});
test('isValidConfigKey still accepts the pre-existing graphify.enabled key', () => {
assert.ok(
isValidConfigKey('graphify.enabled'),
'regression guard: graphify.enabled must remain a valid key',
);
});
test('CANONICAL_CONFIG_DEFAULTS.graphify.auto_update is false', () => {
assert.ok(
CANONICAL_CONFIG_DEFAULTS.graphify !== undefined,
'CANONICAL_CONFIG_DEFAULTS must expose a graphify section',
);
assert.strictEqual(
CANONICAL_CONFIG_DEFAULTS.graphify.auto_update,
false,
'graphify.auto_update default must be false (opt-in per issue #3347 AC)',
);
});
test('config-set graphify.auto_update true succeeds', (t) => {
const tmpDir = createTempProject();
t.after(() => cleanup(tmpDir));
const result = runGsdTools(
['config-set', 'graphify.auto_update', 'true'],
tmpDir,
);
assert.ok(
result.success,
[
'config-set graphify.auto_update true should succeed,',
'got:',
'stdout: ' + result.output,
'stderr: ' + result.error,
].join('\n'),
);
});
test('config-set graphify.auto_update true writes to config.json', (t) => {
const tmpDir = createTempProject();
t.after(() => cleanup(tmpDir));
runGsdTools(['config-set', 'graphify.auto_update', 'true'], tmpDir);
const configPath = path.join(tmpDir, '.planning', 'config.json');
const config = JSON.parse(fs.readFileSync(configPath, 'utf8'));
assert.strictEqual(
config.graphify?.auto_update,
true,
'config.json must have graphify.auto_update: true after config-set',
);
});
test('config-set graphify.auto_update false persists too', (t) => {
const tmpDir = createTempProject();
t.after(() => cleanup(tmpDir));
runGsdTools(['config-set', 'graphify.auto_update', 'true'], tmpDir);
runGsdTools(['config-set', 'graphify.auto_update', 'false'], tmpDir);
const configPath = path.join(tmpDir, '.planning', 'config.json');
const config = JSON.parse(fs.readFileSync(configPath, 'utf8'));
assert.strictEqual(
config.graphify?.auto_update,
false,
'config-set must round-trip true → false',
);
});
test('graphifyStatus folds auto-update status=failed into stale=true', (t) => {
const tmpDir = makeStatusProject({
ts: '2026-05-15T12:00:00Z',
status: 'failed',
exit_code: 1,
duration_ms: 1234,
head_at_build: 'abcdef0',
graphify_version: null,
});
t.after(() => cleanup(tmpDir));
const s = graphifyStatus(tmpDir);
assert.strictEqual(s.stale, true, 'auto-build failure must set stale=true');
assert.ok(s.last_build_auto_update, 'last_build_auto_update must be exposed');
assert.strictEqual(s.last_build_auto_update.status, 'failed');
assert.strictEqual(s.last_build_auto_update.exit_code, 1);
});
test('graphifyStatus folds auto-update status=running into stale=true', (t) => {
const tmpDir = makeStatusProject({
ts: '2026-05-15T12:00:00Z',
status: 'running',
exit_code: null,
duration_ms: null,
head_at_build: 'abcdef0',
graphify_version: null,
});
t.after(() => cleanup(tmpDir));
const s = graphifyStatus(tmpDir);
assert.strictEqual(s.stale, true, 'auto-build in-flight must set stale=true');
assert.strictEqual(s.last_build_auto_update.status, 'running');
});
test('graphifyStatus leaves stale alone when auto-update status=ok and graph is fresh', (t) => {
const tmpDir = makeStatusProject({
ts: '2026-05-15T12:00:00Z',
status: 'ok',
exit_code: 0,
duration_ms: 1234,
head_at_build: 'abcdef0',
graphify_version: null,
});
t.after(() => cleanup(tmpDir));
const s = graphifyStatus(tmpDir);
assert.strictEqual(s.stale, false, 'fresh graph + ok auto-build => not stale');
assert.strictEqual(s.last_build_auto_update.status, 'ok');
});
test('graphifyStatus exposes last_build_auto_update: null when status file absent', (t) => {
const tmpDir = makeStatusProject(null);
t.after(() => cleanup(tmpDir));
const s = graphifyStatus(tmpDir);
assert.strictEqual(s.last_build_auto_update, null);
assert.strictEqual(s.stale, false, 'no status file => stale follows mtime only');
});
test('config-set graphify.auto_update does not perturb sibling graphify.enabled', (t) => {
const tmpDir = createTempProject();
t.after(() => cleanup(tmpDir));
runGsdTools(['config-set', 'graphify.enabled', 'true'], tmpDir);
runGsdTools(['config-set', 'graphify.auto_update', 'true'], tmpDir);
const configPath = path.join(tmpDir, '.planning', 'config.json');
const config = JSON.parse(fs.readFileSync(configPath, 'utf8'));
assert.strictEqual(
config.graphify?.enabled,
true,
'graphify.enabled must be preserved when setting graphify.auto_update',
);
assert.strictEqual(
config.graphify?.auto_update,
true,
'graphify.auto_update must coexist with graphify.enabled',
);
});
});
// ─── hook tests ─────────────────────────────────────────────────────────────
const isWindows = process.platform === 'win32';
const HOOK = path.join(__dirname, '..', 'hooks', 'gsd-graphify-update.sh');
function createTempGitRepo(opts = {}) {
const tmpDir = fs.mkdtempSync(path.join(os.tmpdir(), 'gsd-3347-'));
spawnSync('git', ['init', '-b', opts.defaultBranch || 'main'], {
cwd: tmpDir,
stdio: 'ignore',
});
execFileSync('git', ['config', 'user.email', 'test@example.com'], { cwd: tmpDir });
execFileSync('git', ['config', 'user.name', 'Test'], { cwd: tmpDir });
fs.writeFileSync(path.join(tmpDir, 'README.md'), '# test\n');
execFileSync('git', ['add', 'README.md'], { cwd: tmpDir });
execFileSync('git', ['commit', '-m', 'init'], { cwd: tmpDir, stdio: 'ignore' });
fs.mkdirSync(path.join(tmpDir, '.planning'), { recursive: true });
if (opts.config !== undefined) {
fs.writeFileSync(
path.join(tmpDir, '.planning', 'config.json'),
JSON.stringify(opts.config, null, 2),
);
}
return tmpDir;
}
function makeMockGraphifyBin(tmpDir, { exitCode = 0, sleepMs = 0 } = {}) {
const binDir = path.join(tmpDir, '.mock-bin');
fs.mkdirSync(binDir, { recursive: true });
const script = path.join(binDir, 'graphify');
const body = [
'#!/usr/bin/env bash',
'set -u',
sleepMs ? `sleep ${(sleepMs / 1000).toFixed(3)}` : '',
'mkdir -p graphify-out',
'echo \'{"nodes":[],"edges":[]}\' > graphify-out/graph.json',
'echo "mock report" > graphify-out/GRAPH_REPORT.md',
'echo "<html></html>" > graphify-out/graph.html',
`exit ${exitCode}`,
]
.filter(Boolean)
.join('\n');
fs.writeFileSync(script, body + '\n', { mode: 0o755 });
return binDir;
}
function runHook(tmpDir, toolPayload, { env = {}, pathPrepend = '' } = {}) {
const PATH = pathPrepend
? `${pathPrepend}${path.delimiter}${process.env.PATH || ''}`
: process.env.PATH || '';
// 30000ms: already bounded pre-migration (unchanged) — this is the `slow`
// suite and the hook itself dispatches a detached graphify rebuild that
// some tests wait on separately; the hook's own synchronous return (gate
// checks + status-file write) is fast, so 30s stays generous headroom.
const r = seamRunHook(HOOK, [], {
interpreter: 'bash',
cwd: tmpDir,
input: JSON.stringify(toolPayload),
env: {
...process.env,
PATH,
CI: '',
...env,
},
timeoutMs: 30000,
});
return { status: r.exitCode, stdout: r.stdout, stderr: r.stderr };
}
// Wait until the detached rebuild writes a terminal status, with a generous
// deadline. The detached subprocess can be slow under contended Docker; all
// assertions here are outcome-based, so we wait for the real terminal state
// rather than guessing a tight wall-clock budget (#382).
async function waitForBuildStatus(statusPath, terminal /* e.g. new Set(['ok']) */, { deadlineMs = 30000, stepMs = 25 } = {}) {
const deadline = Date.now() + deadlineMs;
let status;
while (Date.now() < deadline) {
if (fs.existsSync(statusPath)) {
try {
const parsed = JSON.parse(fs.readFileSync(statusPath, 'utf8'));
status = parsed;
if (parsed && terminal.has(parsed.status)) return parsed;
} catch { /* detached writer can briefly expose a partial JSON write */ }
}
await delay(stepMs);
}
return status; // last seen (may be undefined / non-terminal) — assertions below will report
}
async function cleanupHookRepo(tmpDir) {
// The hook detaches a graphify-rebuild subprocess that may still be writing
// into tmpDir when the test body returns. Wait for the rebuild PID to exit
// (lock file disappears on subprocess EXIT trap), then retry rmSync to
// absorb any remaining transient ENOTEMPTY race.
//
// Uses delay() instead of Atomics.wait so we yield the event loop
// without blocking the thread. Budget: 15 s — bumped from 4 s to give
// the detached child more time to exit under contended Docker (#382).
const lockPath = path.join(tmpDir, '.planning/graphs/.rebuild.lock');
const lockDeadline = Date.now() + 15000;
while (Date.now() < lockDeadline) {
if (!fs.existsSync(lockPath)) break;
try {
const pid = parseInt(fs.readFileSync(lockPath, 'utf8'), 10);
if (!Number.isFinite(pid) || pid <= 0) break;
execFileSync('kill', ['-0', String(pid)], { stdio: 'ignore' });
} catch {
break; // PID dead → safe to clean up
}
await delay(50); // yield 50 ms, then re-check
}
try {
// eslint-disable-next-line local/no-raw-rmsync-in-tests -- best-effort teardown: error is swallowed so cleanup() (which propagates) cannot be used here; a residual temp dir after a detached-subprocess race is harmless (#382)
fs.rmSync(tmpDir, { recursive: true, force: true, maxRetries: 8, retryDelay: 100 });
} catch { /* best-effort teardown: a residual temp dir is harmless; never fail the test (#382) */ }
}
describe('hook — bail paths (no side effects)',
{ skip: isWindows ? 'POSIX-only: harness spawns bash + kill -0 + sleep; the hook itself is a bash script under test' : false },
() => {
test('non-Bash tool call exits 0 with no status file', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const r = runHook(tmpDir, { tool_name: 'Edit', tool_input: { file_path: 'x' } });
assert.strictEqual(r.status, 0, 'hook must exit 0 on non-Bash tool');
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
'no status file should be created when bailing',
);
});
test('Bash but non-HEAD-advancing command exits 0 with no status file', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const r = runHook(tmpDir, { tool_name: 'Bash', tool_input: { command: 'ls -la' } });
assert.strictEqual(r.status, 0);
assert.ok(!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')));
});
test('git commit but graphify.enabled=false → no dispatch', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: false, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const r = runHook(tmpDir, { tool_name: 'Bash', tool_input: { command: 'git commit -m x' } });
assert.strictEqual(r.status, 0);
assert.ok(!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')));
});
test('git commit but graphify.auto_update=false → no dispatch (opt-in)', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: false } },
});
t.after(() => cleanupHookRepo(tmpDir));
const r = runHook(tmpDir, { tool_name: 'Bash', tool_input: { command: 'git commit -m x' } });
assert.strictEqual(r.status, 0);
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
'opt-in default-off: auto_update=false must suppress dispatch',
);
});
test('CI=true → no dispatch even with both gates true', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir);
const r = runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ env: { CI: 'true' }, pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
assert.ok(!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')));
});
test('on non-default branch → no dispatch', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
execFileSync('git', ['checkout', '-b', 'worktree-agent-abc'], {
cwd: tmpDir,
stdio: 'ignore',
});
const mockBin = makeMockGraphifyBin(tmpDir);
const r = runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
'branch check must filter worktree-agent-* (non-default-branch) commits',
);
});
test('graphify binary not on PATH → silent exit 0', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
// Note: do NOT prepend mock bin; rely on real PATH not having graphify
const r = runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ env: { PATH: '/usr/bin:/bin' } },
);
assert.strictEqual(r.status, 0, 'must not break commits when graphify missing');
});
});
describe('hook — dispatch path (all gates pass)',
{ skip: isWindows ? 'POSIX-only: harness spawns bash + kill -0 + sleep; the hook itself is a bash script under test' : false },
() => {
test('writes status file with status=running synchronously before returning', async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
// Sleep 2s in mock so we can observe the running state before completion
const mockBin = makeMockGraphifyBin(tmpDir, { sleepMs: 2000 });
const r = runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0, 'hook must return 0');
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
assert.ok(fs.existsSync(statusPath), 'status file must be written synchronously');
const status = JSON.parse(fs.readFileSync(statusPath, 'utf8'));
assert.strictEqual(status.status, 'running', 'initial status must be "running"');
assert.ok(/^[0-9a-f]{7,40}$/.test(status.head_at_build), 'head_at_build must be a commit sha');
// Await the detached rebuild's terminal status so the subprocess has exited
// before the test returns. A dangling detached rebuild can outlive the chunk
// under `node --test --test-force-exit` and surface as a non-zero chunk exit.
const done = await waitForBuildStatus(statusPath, new Set(['ok', 'failed']));
assert.ok(done && (done.status === 'ok' || done.status === 'failed'),
'detached rebuild must reach a terminal status before the test returns');
});
test('completes to status=ok after detached graphify run succeeds', async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir, { exitCode: 0, sleepMs: 200 });
runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
// Wait for the detached rebuild process to write the final status file.
// Uses waitForBuildStatus with a generous 30 s deadline so contended
// Docker is never racing a tight budget (#382). All assertions are
// outcome-based; the deadline is deterministic, not a timing assertion.
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
const status = await waitForBuildStatus(statusPath, new Set(['ok']));
assert.ok(status, 'status file must exist after dispatch');
assert.strictEqual(status.status, 'ok', 'mock graphify exit=0 → status ok');
assert.strictEqual(status.exit_code, 0);
assert.ok(typeof status.duration_ms === 'number' && status.duration_ms >= 0);
});
test('completes to status=failed when graphify exits non-zero', async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir, { exitCode: 1, sleepMs: 100 });
runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
// Wait for the detached rebuild process to write the final status file.
// Uses waitForBuildStatus with a generous 30 s deadline so contended
// Docker is never racing a tight budget (#382). All assertions are
// outcome-based; the deadline is deterministic, not a timing assertion.
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
const status = await waitForBuildStatus(statusPath, new Set(['failed']));
assert.ok(status, 'status file must exist after dispatch');
assert.strictEqual(status.status, 'failed', 'mock graphify exit=1 → status failed');
assert.strictEqual(status.exit_code, 1);
});
test('lock file with a live PID prevents concurrent dispatch', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
fs.mkdirSync(path.join(tmpDir, '.planning/graphs'), { recursive: true });
// Seed a live-PID lock pointing at our own process — kill -0 will succeed
fs.writeFileSync(path.join(tmpDir, '.planning/graphs/.rebuild.lock'), String(process.pid));
const mockBin = makeMockGraphifyBin(tmpDir);
const r = runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
// Status file should NOT be written because a rebuild is in flight
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
'live PID lock must suppress dispatch',
);
});
test('stale lock file (dead PID) is treated as absent', async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
fs.mkdirSync(path.join(tmpDir, '.planning/graphs'), { recursive: true });
// PID 1 is init; kill -0 1 succeeds for root but fails for non-root.
// Use a very large PID number unlikely to exist (max pid = 4194304 on linux).
fs.writeFileSync(path.join(tmpDir, '.planning/graphs/.rebuild.lock'), '4194303');
const mockBin = makeMockGraphifyBin(tmpDir, { sleepMs: 500 });
const r = runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
assert.ok(fs.existsSync(statusPath), 'stale lock must not block dispatch');
// Await terminal status so the detached rebuild is gone before returning (see above).
const done = await waitForBuildStatus(statusPath, new Set(['ok', 'failed']));
assert.ok(done && (done.status === 'ok' || done.status === 'failed'),
'detached rebuild must reach a terminal status before the test returns');
});
test('respects git.base_branch config override (default branch != main)', async (t) => {
const tmpDir = createTempGitRepo({
defaultBranch: 'trunk',
config: {
graphify: { enabled: true, auto_update: true },
git: { base_branch: 'trunk' },
},
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir, { sleepMs: 100 });
const r = runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
assert.ok(
fs.existsSync(statusPath),
'hook must honor git.base_branch when default branch is not main',
);
// Await terminal status so the detached rebuild is gone before returning (see above).
const done = await waitForBuildStatus(statusPath, new Set(['ok', 'failed']));
assert.ok(done && (done.status === 'ok' || done.status === 'failed'),
'detached rebuild must reach a terminal status before the test returns');
});
});
// #2304 — Kimi tool vocabulary engages the hook: Kimi CLI registers this
// hook with matcher 'Shell' and forwards tool_name 'Shell' (possibly
// module-qualified). kimi-cli's Shell.Params names its field `command`
// (src/kimi_cli/tools/shell/__init__.py), same as Claude's Bash, so only
// the tool name needs normalization. Pre-fix, the Gate 1 `== "Bash"` check
// never matched on Kimi and the auto-rebuild was silently dormant.
describe('hook — Kimi tool vocabulary (#2304)',
{ skip: isWindows ? 'POSIX-only: harness spawns bash to invoke the .sh hook under test' : false },
() => {
test('dispatches on Kimi tool_name Shell (all gates pass)', async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir);
const r = runHook(
tmpDir,
{ tool_name: 'Shell', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0, 'hook must return 0');
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
assert.ok(
fs.existsSync(statusPath),
'Kimi Shell must pass Gate 1 and dispatch — pre-fix the hook was silently dormant on Kimi (#2304)',
);
const done = await waitForBuildStatus(statusPath, new Set(['ok', 'failed']));
assert.ok(done && (done.status === 'ok' || done.status === 'failed'),
'detached rebuild must reach a terminal status before the test returns');
});
test('dispatches on module-qualified kimi_cli.tools.shell:Shell', async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir);
const r = runHook(
tmpDir,
{ tool_name: 'kimi_cli.tools.shell:Shell', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
assert.ok(fs.existsSync(statusPath),
'module-qualified Kimi tool name must be recognized after prefix strip');
const done = await waitForBuildStatus(statusPath, new Set(['ok', 'failed']));
assert.ok(done && (done.status === 'ok' || done.status === 'failed'),
'detached rebuild must reach a terminal status before the test returns');
});
test('negative control: Kimi WriteFile does NOT dispatch', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir);
const r = runHook(
tmpDir,
// A command-shaped payload under a non-Shell Kimi tool: the name
// normalization must not widen Gate 1 beyond Shell→Bash.
{ tool_name: 'kimi_cli.tools.file:WriteFile', tool_input: { command: 'git commit -m x' } },
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
'non-Shell Kimi tools must still bail at Gate 1',
);
});
});
describe('hook — HEAD-advancing command matchers',
{ skip: isWindows ? 'POSIX-only: harness spawns bash to invoke the .sh hook under test' : false },
() => {
for (const cmd of [
'git commit -m fix',
'git merge feature',
'git pull --ff-only',
'git rebase --continue',
'git cherry-pick abc123',
// #3653 — `gsd-tools query commit` invokes git via spawnSync('git', [...]),
// so the substring "git commit" never appears in tool_input.command.
// The hook must match the user-facing SDK invocation directly.
'gsd-tools query commit "docs: probe" --files .planning/STATE.md',
'npx gsd-tools query commit "docs: probe" --files .planning/STATE.md',
]) {
test(`dispatches on: ${cmd}`, async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir, { sleepMs: 100 });
runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: cmd } },
{ pathPrepend: mockBin },
);
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
// Wait for any terminal status; we only assert the file exists,
// not which terminal value it holds (#382: generous deadline).
await waitForBuildStatus(statusPath, new Set(['ok', 'failed']));
assert.ok(
fs.existsSync(statusPath),
`must dispatch for HEAD-advancing op: ${cmd}`,
);
});
}
test('does NOT dispatch on SDK commit-to-subrepo prefix collision', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir, { sleepMs: 100 });
const r = runHook(
tmpDir,
{
tool_name: 'Bash',
tool_input: {
command: 'gsd-tools query commit-to-subrepo "msg" --files packages/foo',
},
},
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
'must NOT dispatch for commit-to-subrepo, which does not advance the outer repo HEAD',
);
});
// #1772 — agent runtimes (Claude Code's Bash tool among them) routinely
// emit HEAD-advancing commits as multi-line scripts (`cd /path` then
// `git add` then `git commit …`). The hook joins tool_name + "\n" +
// tool_input.command and must match the command across ALL its lines
// (line 2 through EOF), not just line 2 — otherwise a `git commit` that
// is not on the first command line silently no-ops the rebuild.
for (const cmd of [
"cd /tmp/repo\ngit add .\ngit commit -m 'multi-line commit'",
"cd /tmp/repo\ngit merge feature-branch",
"git fetch origin\ngit pull --ff-only",
]) {
test(`dispatches on multi-line command (#1772): ${cmd.split('\n').slice(0, 2).join(' ⏎ ')}…`, async (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir, { sleepMs: 100 });
runHook(
tmpDir,
{ tool_name: 'Bash', tool_input: { command: cmd } },
{ pathPrepend: mockBin },
);
const statusPath = path.join(tmpDir, '.planning/graphs/.last-build-status.json');
await waitForBuildStatus(statusPath, new Set(['ok', 'failed']));
assert.ok(
fs.existsSync(statusPath),
`must dispatch for multi-line command where the HEAD-advancing op is not on line 1 (#1772): ${cmd}`,
);
});
}
test('multi-line command with NO HEAD-advancing op still no-ops (#1772 no-regression)', (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const mockBin = makeMockGraphifyBin(tmpDir, { sleepMs: 100 });
const r = runHook(
tmpDir,
{
tool_name: 'Bash',
tool_input: {
// Multi-line, but only non-HEAD-advancing ops. Widening line
// extraction to 2..EOF must not cause a spurious dispatch.
command: 'cd /tmp/repo\nls -la\necho done',
},
},
{ pathPrepend: mockBin },
);
assert.strictEqual(r.status, 0);
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
'multi-line command without a HEAD-advancing git op must still no-op (#1772)',
);
});
// #3653 — only the SDK `commit` verb invokes git internally. Other
// `gsd-tools query` verbs (phase.complete, roadmap.update-plan-progress,
// state.begin-phase) mutate .md files but do NOT advance HEAD; matching
// them would cause a spurious rebuild per state mutation.
for (const cmd of [
'gsd-tools query phase.complete 109',
'gsd-tools query roadmap.update-plan-progress 109 W001',
'gsd-tools query state.begin-phase 110',
]) {
test(`does NOT dispatch on non-HEAD-advancing SDK verb: ${cmd}`, (t) => {
const tmpDir = createTempGitRepo({
config: { graphify: { enabled: true, auto_update: true } },
});
t.after(() => cleanupHookRepo(tmpDir));
const r = runHook(tmpDir, { tool_name: 'Bash', tool_input: { command: cmd } });
assert.strictEqual(r.status, 0);
assert.ok(
!fs.existsSync(path.join(tmpDir, '.planning/graphs/.last-build-status.json')),
`must NOT dispatch for non-HEAD-advancing SDK verb: ${cmd}`,
);
});
}
});
});