The windows-test-parity ratchet greps test source for fs.rmSync-without-
maxRetries (and six other Windows-portability anti-patterns), failing when an
integer offender COUNT exceeds a frozen baseline (rmSync: 95). A count ratchet
is a Goodhart metric: fixing one offender and adding another keeps the count
constant, so a new defect slips through green. Replace it — and every other
count ratchet in the repo — with a layered, masking-proof design.
Behavioral seam test
- tests/helpers-cleanup.test.cjs proves helpers.cleanup() carries the Windows
EBUSY retry budget. cleanup() delegates retries to Node's fs.rmSync via
maxRetries (it owns no loop), so the test asserts the option contract
(recursive/force/maxRetries>0/retryDelay>0) + real-FS removal + the cwd-guard,
rather than a loop that does not exist. The EBUSY risk is now tested ONCE at
the helper, not approximated textually at every call site.
Write-time ESLint rule (AST-accurate, replaces the grep)
- eslint-rules/no-raw-rmsync-in-tests.cjs (error in tests/**/*.test.cjs) bans
raw fs.rmSync, steering to cleanup(). Catches member, computed (fs['rmSync']),
destructured and aliased forms; escape hatch is inline
`// eslint-disable-next-line local/no-raw-rmsync-in-tests -- <reason>` only.
- Migrated 336 raw fs.rmSync teardown calls across ~116 test files to cleanup().
~18 genuinely load-bearing sites (mid-test SUT/fault-injection removals,
error-swallowing or name-colliding local teardown helpers) keep the raw call
with an inline eslint-disable + reason.
Shared anti-ratchet primitive
- scripts/lib/allowlist-ratchet.cjs:
- assertWithinAllowlist: fails on NOVEL ids (new offender introduced) AND on
STALE ids (a known offender was fixed but not pruned) — identity, not count,
and a ratchet DOWN toward zero.
- assertTightCeiling: a size/length budget whose ceiling must stay within a
grace band of the high-water mark, so budgets may only tighten, never creep.
Ratchets converted onto the primitive
- windows-test-parity-guard.test.cjs: rmSync rule deleted (now ESLint-enforced);
the remaining six patterns moved from integer baselines to named-set
allowlists with ratchet-down.
- scripts/lint-test-file-count.{cjs,allowlist.json}: per-module integer counts →
named filename sets (closes the swap-a-file-keep-the-count blind spot); a
module dropping under cap now FAILS to force pruning its allowlist entry.
- enh-2790 skill-count `<= 63` → named skill allowlist (ratchets toward ~58).
Size budgets hardened (tighten-only)
- agent-size / workflow-size / feat-3039 help-tiered: ceilings lowered to the
current high-water mark and an assertTightCeiling anti-creep check added per
tier. Fixed external-contract limits (description ≤100 chars, agent ≤100 KB)
are intentionally left as-is — they are not grandfathered creeping budgets.
No user-facing behavior change (tests + tooling only); no USER_FACING_PREFIXES
touched, so no changeset fragment is required.
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
161 lines
6.1 KiB
JavaScript
161 lines
6.1 KiB
JavaScript
/**
|
|
* Regression tests for issue #2256 — per-agent model_overrides transport
|
|
* for Codex and OpenCode runtimes.
|
|
*
|
|
* The bug: model_overrides set in per-project `.planning/config.json` were
|
|
* never read by the Codex / OpenCode install paths, which only probed
|
|
* `~/.gsd/defaults.json`. As a result, the configured per-agent model was
|
|
* dropped and child agents inherited the runtime's default model.
|
|
*
|
|
* These tests lock in the fix: per-project overrides must be honored, and
|
|
* per-project keys must win over global when both are present.
|
|
*/
|
|
|
|
process.env.GSD_TEST_MODE = '1';
|
|
|
|
const { test, describe, beforeEach, afterEach } = require('node:test');
|
|
const assert = require('node:assert/strict');
|
|
const fs = require('fs');
|
|
const path = require('path');
|
|
const os = require('os');
|
|
|
|
const isWindows = process.platform === 'win32';
|
|
|
|
const {
|
|
readGsdEffectiveModelOverrides,
|
|
generateCodexAgentToml,
|
|
convertClaudeToOpencodeFrontmatter,
|
|
getCodexSkillAdapterHeader,
|
|
} = require('../bin/install.js');
|
|
|
|
const { createTempDir, cleanup } = require('./helpers.cjs');
|
|
const makeTmp = (prefix) => createTempDir(`gsd-2256-${prefix}-`);
|
|
|
|
function writeJson(p, obj) {
|
|
fs.mkdirSync(path.dirname(p), { recursive: true });
|
|
fs.writeFileSync(p, JSON.stringify(obj, null, 2));
|
|
}
|
|
|
|
describe('bug #2256 — readGsdEffectiveModelOverrides', () => {
|
|
let projectDir;
|
|
let homeDir;
|
|
let origHome;
|
|
let origUserProfile;
|
|
|
|
beforeEach(() => {
|
|
projectDir = makeTmp('proj');
|
|
homeDir = makeTmp('home');
|
|
origHome = process.env.HOME;
|
|
// On Windows, os.homedir() reads USERPROFILE (not HOME). Tests that
|
|
// need to redirect ~ must override both — otherwise the SUT reads
|
|
// the real user's home and the fixture is invisible.
|
|
origUserProfile = process.env.USERPROFILE;
|
|
process.env.HOME = homeDir;
|
|
if (isWindows) process.env.USERPROFILE = homeDir;
|
|
});
|
|
|
|
afterEach(() => {
|
|
if (origHome === undefined) delete process.env.HOME;
|
|
else process.env.HOME = origHome;
|
|
if (isWindows) {
|
|
if (origUserProfile === undefined) delete process.env.USERPROFILE;
|
|
else process.env.USERPROFILE = origUserProfile;
|
|
}
|
|
cleanup(projectDir);
|
|
cleanup(homeDir);
|
|
});
|
|
|
|
test('returns null when neither source defines model_overrides', () => {
|
|
const result = readGsdEffectiveModelOverrides(projectDir);
|
|
assert.strictEqual(result, null);
|
|
});
|
|
|
|
test('reads overrides from ~/.gsd/defaults.json (global only)', () => {
|
|
writeJson(path.join(homeDir, '.gsd', 'defaults.json'), {
|
|
model_overrides: { 'gsd-codebase-mapper': 'gpt-5-mini' },
|
|
});
|
|
const result = readGsdEffectiveModelOverrides(projectDir);
|
|
assert.deepStrictEqual(result, { 'gsd-codebase-mapper': 'gpt-5-mini' });
|
|
});
|
|
|
|
test('reads overrides from per-project .planning/config.json', () => {
|
|
writeJson(path.join(projectDir, '.planning', 'config.json'), {
|
|
model_overrides: { 'gsd-codebase-mapper': 'claude-haiku-4-5' },
|
|
});
|
|
const result = readGsdEffectiveModelOverrides(projectDir);
|
|
assert.deepStrictEqual(result, { 'gsd-codebase-mapper': 'claude-haiku-4-5' });
|
|
});
|
|
|
|
test('per-project overrides win over global on conflict', () => {
|
|
writeJson(path.join(homeDir, '.gsd', 'defaults.json'), {
|
|
model_overrides: { 'gsd-codebase-mapper': 'global-model', 'gsd-planner': 'opus' },
|
|
});
|
|
writeJson(path.join(projectDir, '.planning', 'config.json'), {
|
|
model_overrides: { 'gsd-codebase-mapper': 'project-model' },
|
|
});
|
|
const result = readGsdEffectiveModelOverrides(projectDir);
|
|
// Per-project wins on conflict; non-conflicting global keys are preserved.
|
|
assert.deepStrictEqual(result, {
|
|
'gsd-codebase-mapper': 'project-model',
|
|
'gsd-planner': 'opus',
|
|
});
|
|
});
|
|
|
|
test('walks up from nested targetDir to find .planning/', () => {
|
|
writeJson(path.join(projectDir, '.planning', 'config.json'), {
|
|
model_overrides: { 'gsd-planner': 'project-opus' },
|
|
});
|
|
const nested = path.join(projectDir, '.codex');
|
|
fs.mkdirSync(nested, { recursive: true });
|
|
const result = readGsdEffectiveModelOverrides(nested);
|
|
assert.deepStrictEqual(result, { 'gsd-planner': 'project-opus' });
|
|
});
|
|
});
|
|
|
|
describe('bug #2256 — Codex adapter embeds per-project override', () => {
|
|
const agentContent = `---\nname: gsd-codebase-mapper\ndescription: Maps codebase\n---\n\nbody\n`;
|
|
|
|
test('generateCodexAgentToml embeds model when override provided', () => {
|
|
const toml = generateCodexAgentToml(
|
|
'gsd-codebase-mapper',
|
|
agentContent,
|
|
{ 'gsd-codebase-mapper': 'gpt-5-mini' },
|
|
);
|
|
assert.match(toml, /^model = "gpt-5-mini"$/m);
|
|
});
|
|
|
|
test('generateCodexAgentToml omits model when no override', () => {
|
|
const toml = generateCodexAgentToml('gsd-codebase-mapper', agentContent, null);
|
|
assert.doesNotMatch(toml, /^model\s*=/m);
|
|
});
|
|
});
|
|
|
|
describe('bug #2256 — OpenCode adapter embeds per-project override', () => {
|
|
test('convertClaudeToOpencodeFrontmatter embeds model on agent frontmatter', () => {
|
|
const input = `---\nname: gsd-codebase-mapper\ndescription: Maps codebase\n---\n\nbody\n`;
|
|
const out = convertClaudeToOpencodeFrontmatter(input, {
|
|
isAgent: true,
|
|
modelOverride: 'claude-haiku-4-5',
|
|
});
|
|
assert.match(out, /^model: claude-haiku-4-5$/m);
|
|
assert.match(out, /^mode: subagent$/m);
|
|
});
|
|
|
|
test('convertClaudeToOpencodeFrontmatter omits model when override absent', () => {
|
|
const input = `---\nname: gsd-codebase-mapper\ndescription: Maps codebase\n---\n\nbody\n`;
|
|
const out = convertClaudeToOpencodeFrontmatter(input, { isAgent: true, modelOverride: null });
|
|
assert.doesNotMatch(out, /^model:/m);
|
|
});
|
|
});
|
|
|
|
describe('bug #2256 — Codex skill adapter header documents transport', () => {
|
|
test('Task(model=...) line no longer says "omit" without explanation', () => {
|
|
const header = getCodexSkillAdapterHeader('gsd-plan-phase');
|
|
// Header must mention that per-agent model_overrides are embedded in agent
|
|
// TOML so spawn_agent picks them up automatically — the old text said
|
|
// "Codex uses per-role config, not inline model selection" which left
|
|
// users thinking their model_overrides were silently ignored.
|
|
assert.match(header, /model_overrides/);
|
|
});
|
|
});
|