cmdAgentSkills' plain (non-JSON) path wrote the <agent_skills> block with
process.stdout.write then immediately called process.exit(0). stdout.write is
async on pipes/files, so on Windows the process tore down before Node flushed
the buffer — the workflows' `$(gsd_run query agent-skills <type>)` capture
received 0 bytes and every ${AGENT_SKILLS_*} substitution expanded empty,
silently dropping configured per-agent skills.
Route the plain path through the existing synchronous output() helper
(writeAllSync, src/io.cts) + return — the same flush-safe mechanism the --json
branch already uses — instead of write + process.exit(0).
Adds a #1400 regression block to tests/agent-skills.test.cjs that captures
stdout via a real file descriptor and asserts the block is non-empty and
byte-identical to the --json .block.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Narrows P3 of ADR-1411 (Resolution Provenance, epic #1411) based on an
adversarial fit-analysis that showed a single Resolution<T> envelope adopted
by agent-skills, capability-state, and capability-writer fails the deletion
test: configured/reason are meaningless for capability verbs, and
capability-writer's errors[] (operation-not-applied) cannot fold into
warnings[]. The only genuinely shared seam is warnings: string[].
Changes:
- src/resolution.cts: new pure types+builder leaf — exports Resolution<T>
{value, configured, reason, warnings}, makeResolution<T>() builder, and
AgentSkillsValue {block, skills_count}. No other src/ imports.
- src/init.cts: cmdAgentSkills --json IR gains additive value:{block,
skills_count} field (built via makeResolution). All existing flat fields
(agent_type, block, skills_count, warnings, configured, reason, source,
degraded) are retained unchanged for back-compat.
- src/capability-state.cts: doc comment on ResolveCapabilityRuntimeStateResult
naming it the canonical read-verb envelope. No JSON change.
- src/capability-writer.cts: doc comment on SetCapabilityStateResult naming it
the canonical mutation-verb result (warnings=advisory, errors=operation-
not-applied). No JSON change.
- CONTEXT.md: new ### Resolution Convention glossary entry after
### Resolution Provenance.
- docs/adr/1411-resolution-provenance.md: P3 narrowing amendment appended.
- tests/resolution.test.cjs: 9 unit tests for makeResolution (new).
- tests/agent-skills.test.cjs: 2 P3 tests for value.block/value.skills_count
and back-compat of all flat fields.
All 277 tests pass (5 suites). npm run lint clean. All lint checks pass.
Part of #1411
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
The empty-IR test asserted `parsed === '' || typeof parsed === 'object'`,
which always passes (typeof null === 'object'). Pin the real contract:
no agent type → `output('', raw, '')` → the --json IR is the empty string.
The nonexistent-skill-path test asserted only `block === ''`. Since #1376
added a warnings[] field to the --json IR, also assert warnings[] names the
skipped path so the test guards the silent-drop regression it is named for.
Test-only; no product behavior change.
Closes#1377
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
* fix(#1374): surface diagnostic when configured agent skills all fail to resolve
buildAgentSkillsBlock returned '' (only ad-hoc per-path stderr warnings) when an agent configured via agent_skills had paths that all failed to resolve — missing SKILL.md, unsafe path, invalid global name, OR a malformed (non-string/non-array) value. query agent-skills --json reported skills_count>0 with an empty block and no machine-readable signal, so a fully-dropped configuration was indistinguishable from a resolved one.
Thread an optional diagnostics collector through buildAgentSkillsBlock: route every skip warning through a warn() helper (stderr + collector), flag truthy-but-malformed config values, emit an aggregate warning when configured paths resolve to zero skills, and surface the collected reasons in a new warnings[] field on the query agent-skills --json IR. Empty arrays and falsy values stay silent (skills_count is honestly 0). skills_count semantics unchanged. Docs updated for the new IR field.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* chore(#1374): backfill changeset PR number (#1376)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
* feat(#1243): consume Claude plugin-provided skills via native Skill-tool directive + grant Skill to agent_skills-consumer agents
- Relax global skill name validation to accept namespaced form `^[A-Za-z0-9_-]+(:[A-Za-z0-9_-]+)*$`
- Namespaced names (containing colon) on claude runtime emit a Skill-tool load directive instead of a @-include line
- Namespaced names on non-claude runtimes are skipped with a warning
- Bare unresolved names retain existing warn-and-skip behavior (no promotion to directive)
- Grant `Skill` tool to all 22 agent_skills consumer agents; 5 generated agents updated via research-profiles.cjs + regen, 17 hand-authored agents edited directly
- Add 16 TDD tests in describe('bug #1243') covering happy/mixed/precedence/negative/cross-runtime/regression/grant cases
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* docs(#1243): document plugin-provided skills in agent_skills
Update the Agent Skills Injection reference in CONFIGURATION.md with
the three entry forms (project-relative, global:<name>,
global:<plugin>:<skill>), the Claude-only runtime behaviour of the
namespaced form and the warn-skip on other runtimes, the plugin
pre-install prerequisite, and the consumer-agent Skill tool grant.
Add docs/how-to/attach-a-plugin-skill-to-a-gsd-agent.md with a
step-by-step guide for installing the plugin, locating the namespaced
skill name, wiring it into agent_skills, and verifying injection.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* fix(#1243): align agent_skills docs with emitted block format + mixed-block regression test (code-review)
- Replace two-section mixed-block example (bogus "Load these plugin-provided skills using the Skill tool:" header) with the actual single-section inline format in CONFIGURATION.md and docs/how-to/attach-a-plugin-skill-to-a-gsd-agent.md
- Fix quoted warning text in how-to doc to exactly match the emitted string: [agent-skills] WARNING: Plugin-namespaced skill "global:<name>" requires a Skill-tool-capable runtime (claude) — skipping on runtime "<runtime>"
- Replace phantom agent slugs (gsd-checker, gsd-researcher, gsd-advisor, gsd-synthesizer) in CONFIGURATION.md Supported Agent Types with real agents/gsd-*.md examples (gsd-plan-checker, gsd-phase-researcher, gsd-code-reviewer, gsd-ui-auditor, gsd-research-synthesizer)
- Add byte-identical mixed-block regression test: one path-resolvable global skill + one plugin-namespaced skill on claude runtime → asserts r.ir.block === single-section interleaved block, no secondary header
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* chore(#1243): regenerate agent-size baseline for the Skill-tool grant
The 22 agent_skills-consumer agents each grew +7 bytes from adding `Skill`
to their tools list; refresh the committed per-agent size baseline (#1074 guard).
* chore(#1243): add Added changeset fragment
* fix(#1243): traceable allow-test-rule ref + separator-agnostic byte-identical tests (CI)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
* feat(#52): add agent_skills_security.trusted_global_roots allowlist
Opt-in allowlist so a global: agent skill whose SKILL.md realpath resolves
outside the default global skills base (e.g. ~/.claude/skills) is accepted
when its real target lies under a user-declared trusted root. Default [] is
byte-identical to prior behavior; the symlink-escape guard is preserved and
simply re-applied against each declared root.
- src/security.cts: loadTrustedGlobalRoots — tilde-expand (~ and ~/), reject
project-relative and dangerously broad roots (filesystem/UNC root, homedir),
realpath-canonicalize each root every run and drop non-existent ones.
- src/init.cts: on base-check failure the guard consults the trusted roots
(hoisted out of the loop); emits a stderr NOTE when a skill is accepted via
a trusted root so the widened boundary is visible.
- src/core.cts: thread agent_skills_security through loadConfig.
- config-schema.manifest.json: allow the new key path.
- docs/CONFIGURATION.md: document the option and its security model.
- tests/agent-skills.test.cjs: unit + end-to-end CLI coverage (regression,
feature, negative, broad-root hardening, stderr NOTE).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* docs(#52): add changeset fragment for trusted_global_roots (#754)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
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>
* refactor(#455): implement typed surfaces to retire grep tests
Production surfaces added:
- hooks/managed-hooks-registry.cjs: new CJS module exporting MANAGED_HOOKS
as a typed array; gsd-check-update-worker.js now requires it instead of
declaring an inline array
- bin/install.js: elevate inline gsdHooks to module-level GSD_UNINSTALL_HOOKS,
export it alongside runtimeMap/allRuntimes (already exported)
- scripts/build-hooks.js: export HOOKS_TO_COPY; guard build() behind
require.main===module so tests can require the file without triggering a build
- get-shit-done/bin/lib/init.cjs: add --json mode to agent-skills command,
emitting typed IR { agent_type, block, skills_count } for test assertions
- get-shit-done/bin/gsd-tools.cjs: wire --json flag for agent-skills dispatch
Category-B source-grep migrations:
- tests/managed-hooks.test.cjs: require MANAGED_HOOKS from registry, drop fs.readFileSync+regex
- tests/orphaned-hooks.test.cjs: require MANAGED_HOOKS+HOOKS_TO_COPY as typed exports
- tests/hooks-opt-in.test.cjs: replace gsdHooks regex-parse with GSD_UNINSTALL_HOOKS import
- tests/install-minimal-hooks.test.cjs: replace gsdHooks regex-parse with GSD_UNINSTALL_HOOKS
- tests/copilot-install.test.cjs: replace src.includes() checks with typed
assertions on runtimeMap, allRuntimes, parseRuntimeInput, buildRuntimePromptText
- tests/agent-skills.test.cjs: migrate to --json typed IR assertions
pending-migration-to-typed-ir token cleared (87 of 87 files):
- 78 files already had source-text-is-the-product; removed duplicate token
- 5 files already used typed assertions; reclassified or annotated
- 4 files required individual reclassification to source-text-is-the-product
or architectural-invariant
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(#455): update workflow-guard test to typed GSD_UNINSTALL_HOOKS import; isolate HOME in runtime-launcher (D) test
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(#455): guard install.js main() behind require.main===module so the typed export is require-safe
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* docs(#455): document --json typed surfaces for agent-skills, progress, validate context
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* docs(#455): add changeset fragment for new --json surfaces
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(#455): complete grep migration for files flagged by lint-tests
The branch commit 4e630d99 stripped `allow-test-rule: pending-migration-to-typed-ir`
from ~80 test files without replacing their assertions or adding the correct
exemption annotation. The files were NOT source-grep tests — they read .md
workflow/agent/command/reference files (source-text-is-the-product) or hook
source files for structural invariants (structural-regression-guard). No
assertion logic was changed; only the correct allow-test-rule annotation was
added to each file per CONTRIBUTING.md exception matrix.
73 files: `source-text-is-the-product` — workflow/agent/command/reference .md
7 files: `structural-regression-guard` — hook .js / bin/install.js structural checks
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: CI Rebase Check <ci@gsd-redux>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Review feedback from @trek-e — three blocking issues and one style fix:
1. **Symlink escape guard** — Added validatePath() call on the resolved
global skill path with allowAbsolute: true. This routes the path
through the existing symlink-resolution and containment logic in
security.cjs, preventing a skill directory symlinked to an arbitrary
location from being injected. The name regex alone prevented
traversal in the literal name but not in the underlying directory.
2. **5 new tests** covering the global: code path:
- global:valid-skill resolves and appears in output
- global:invalid!name rejected by regex, skipped without crash
- global:missing-skill (directory absent) skipped gracefully
- Mix of global: and project-relative paths both resolve
- global: with empty name produces clear warning and skips
3. **Explicit empty-name guard** — Added before the regex check so
"global:" produces "empty skill name" instead of the confusing
'Invalid global skill name ""'.
4. **Style fix** — Hoisted require('os') and globalSkillsBase
calculation out of the loop, alongside the existing validatePath
import at the top of buildAgentSkillsBlock.
All 16 agent-skills tests pass.
* refactor(tests): standardize to node:assert/strict and t.after() per CONTRIBUTING.md
- Replace require('node:assert') with require('node:assert/strict') across
all 73 test files to enforce strict equality (no type coercion)
- Replace try/finally cleanup blocks with t.after() hooks in core.test.cjs
and hooks-opt-in.test.cjs per the test lifecycle standards
- Utility functions in codex-config and security-scan retain try/finally
as that is appropriate for per-function resource guards, not lifecycle hooks
Closes#1674
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* perf(tests): add --test-concurrency=4 to test runner for parallel file execution
Node.js --test-concurrency controls how many test files run as parallel child
processes. Set to 4 by default, configurable via TEST_CONCURRENCY env var.
Fixes tests at a known level rather than inheriting os.availableParallelism()
which varies across CI environments.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* fix(security): allowlist verify.test.cjs in prompt-injection scanner
tests/verify.test.cjs uses <human>...</human> as GSD phase task-type
XML (meaning "a human should verify this step"), which matches the
scanner's fake-message-boundary pattern for LLM APIs. This is a
false positive — add it to the allowlist alongside the other test files
that legitimately contain injection-adjacent patterns.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Add agent_skills config section that maps agent types to skill directory
paths. At spawn time, workflows load configured skills and inject them
as <agent_skills> blocks in Task() prompts, giving subagents access to
project-specific skill files.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
