Codex compatibility note:
- Invoke repository skills with
$skill-name in Codex; this mirrored copy rewrites legacy Claude /skill-name references.
- Task tracker mandate: BEFORE executing any workflow or skill step, create/update task tracking for all steps and keep it synchronized as progress changes.
- User-question prompts mean to ask the user directly in Codex.
- Ignore Claude-specific mode-switch instructions when they appear.
- Strict execution contract: when a user explicitly invokes a skill, execute that skill protocol as written.
- Subagent authorization: when a skill is user-invoked or AI-detected and its protocol requires subagents, that skill activation authorizes use of the required
spawn_agent subagent(s) for that task.
- Do not skip, reorder, or merge protocol steps unless the user explicitly approves the deviation first.
- For workflow skills, execute each listed child-skill step explicitly and report step-by-step evidence.
- If a required step/tool cannot run in this environment, stop and ask the user before adapting.
Codex Project-Reference Loading (No Hooks)
Codex uses static project-reference loading instead of runtime-injected project docs.
When coding, planning, debugging, testing, or reviewing, open project docs explicitly using this routing.
Always read:
docs/project-config.json (project-specific paths, commands, modules, and workflow/test settings)docs/project-reference/docs-index-reference.md (routes to the full docs/project-reference/* catalog)docs/project-reference/lessons.md (always-on guardrails and anti-patterns)
Missing/stale context route: If docs/project-config.json, the docs index, lessons.md, CLAUDE.md, AGENTS.md, or any task-required reference doc is missing or stale, auto-run $project-init or the narrow setup route ($project-config, $docs-init, $scan-all, $scan --target=<key>, $claude-md-init) before ordinary project-specific work. If Codex mirrors or AGENTS.md are missing/stale, ask the user to run $sync-codex; do not auto-run it.
Situation-based docs:
- Project structure/architecture/tech-stack/deployment/setup (any layer — backend, frontend, or infra):
project-structure-reference.md
- Backend/CQRS/API/domain/entity changes:
backend-patterns-reference.md, domain-entities-reference.md
- Frontend/UI/styling/design-system:
frontend-patterns-reference.md, scss-styling-guide.md, design-system/README.md
- Spec authoring,
docs/specs/ pathing, or TC format: feature-spec-reference.md, spec-system-reference.md, spec-principles.md
- Behavior/public-contract changes or spec-test-code sync:
workflow-spec-test-code-cycle-reference.md plus the spec docs above
- Derived spec indexes/ERDs/reimplementation guides:
spec-system-reference.md and source Feature Specs under docs/specs/
- Integration test implementation/review:
integration-test-reference.md
- E2E test implementation/review:
e2e-test-reference.md
- Code review/audit work:
code-review-rules.md plus domain docs above based on changed files
Do not read all docs blindly. Start from docs-index-reference.md, then open only relevant files for the task.
Quick Summary
Goal: Generate professional release notes from git commits with automated categorization, service detection, and validation.
Workflow:
- Parse Commits —
parse-commits.cjs <base> <head> extracts structured data from git
- Categorize — Pipe through
categorize-commits.cjs for user-facing vs internal sections
- Render —
render-template.cjs --version vX.Y.Z generates markdown with Summary, What's New, Improvements, Bug Fixes, Breaking Changes, Technical Details
Key Rules:
- Pipeline: parse → categorize → render → validate → transform
- Advanced: Service detection, breaking change analysis, PR metadata, contributor stats, version bumping
- Human Review: Generated notes are Draft status, require review/enhance/approve before publish
- Validation:
validate-notes.cjs scores against quality rules (100 points)
Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).
Release Notes Generation Skill
Generate professional release notes from git commits between two refs with automated categorization.
Invocation
$release-notes [base] [head] [--version vX.Y.Z] [--output path]
Examples:
# Generate release notes for commits since last tag
$release-notes v1.0.0 HEAD --version v1.1.0
# Compare branches
$release-notes main feature/new-auth --version v2.0.0-beta
# Output to specific file
$release-notes v1.0.0 HEAD --version v1.1.0 --output docs/release-notes/250111-v1.1.0.md
Workflow
Step 1: Parse Commits
Execute the commit parser to extract structured data from git history:
node .claude/skills/release-notes/lib/parse-commits.cjs <base> <head> [--with-files]
Output: JSON with commits array containing:
hash, shortHash - Commit identifierstype, scope, description - Conventional commit partsbreaking - Boolean for breaking changesauthor, date - Attributionfiles - Changed files (with --with-files flag)
Step 2: Categorize Commits
Pipe parsed commits through the categorizer:
node .claude/skills/release-notes/lib/parse-commits.cjs <base> <head> | \
node .claude/skills/release-notes/lib/categorize-commits.cjs
Categorization Rules:
| Type |
Category |
User-Facing |
feat |
features |
Yes |
fix |
fixes |
Yes |
perf |
improvements |
Yes |
docs |
docs |
Yes (unless internal) |
refactor |
improvements |
Technical only |
test, ci, build, chore, style |
internal |
No |
Excluded Patterns:
chore(deps): - Dependency updateschore(config): - Configuration changes[skip changelog] - Explicit skip[ci skip] - CI markers
Step 3: Render Markdown
Generate the final release notes document:
node .claude/skills/release-notes/lib/parse-commits.cjs <base> <head> | \
node .claude/skills/release-notes/lib/categorize-commits.cjs | \
node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0 --output docs/release-notes/250111-v1.1.0.md
Complete Pipeline
For generating release notes in a single command:
# Full pipeline with output to file
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
node .claude/skills/release-notes/lib/categorize-commits.cjs | \
node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0 --output docs/release-notes/250111-v1.1.0.md
# Pipeline to stdout for review
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
node .claude/skills/release-notes/lib/categorize-commits.cjs | \
node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0
Advanced Features
Service Boundary Detection
Analyze which services are affected by the release:
# Parse with file changes, then detect services
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD --with-files | \
node .claude/skills/release-notes/lib/detect-services.cjs
Output: Service impact analysis with severity levels (critical, high, medium, low)
Breaking Change Analysis
Enhanced breaking change detection with migration info extraction:
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
node .claude/skills/release-notes/lib/categorize-commits.cjs | \
node .claude/skills/release-notes/lib/detect-breaking.cjs
Detects:
BREAKING CHANGE: in commit body! suffix on commit type (e.g., feat!:)- Migration instructions
PR Metadata Extraction
Extract and link pull request information:
# Extract PR numbers from commit messages
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
node .claude/skills/release-notes/lib/extract-pr-metadata.cjs
# With GitHub API enrichment (requires gh CLI)
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
node .claude/skills/release-notes/lib/extract-pr-metadata.cjs --fetch-gh
Extracts: PR numbers, titles, labels, authors from commits
Contributor Statistics
Generate detailed contributor stats:
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
node .claude/skills/release-notes/lib/contributor-stats.cjs
Output: Contributor list with commit counts, feature/fix breakdown
Version Bumping
Automatically determine and bump semantic version based on commit types:
# Auto-bump based on commits (feat→minor, fix→patch, BREAKING→major)
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
node .claude/skills/release-notes/lib/bump-version.cjs
# Bump with prerelease tag
node .claude/skills/release-notes/lib/bump-version.cjs --prerelease beta
# Per-service versioning
node .claude/skills/release-notes/lib/bump-version.cjs --service {service-name}
# Dry run (don't write version file)
node .claude/skills/release-notes/lib/bump-version.cjs --dry-run
Version Files:
- Root:
.version
- Per-service:
.versions/<service-name>.version
Quality Validation
Validate release notes against quality rules:
# Validate with default threshold (70)
node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md
# Custom threshold
node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md --threshold 80
# JSON output for CI
node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md --json
Validation Rules (100 points total):
| Rule |
Weight |
Description |
| summary_exists |
15 |
Has Summary section |
| summary_not_empty |
10 |
Summary has content |
| has_version |
10 |
Version number present |
| features_documented |
10 |
Features properly formatted |
| fixes_documented |
10 |
Bug fixes properly formatted |
| no_broken_links |
10 |
No empty link references |
| contributors_listed |
10 |
Contributors section present |
| has_date |
5 |
Date present |
| no_todo_markers |
5 |
No TODO/FIXME markers |
| proper_heading_hierarchy |
5 |
Proper H1→H2 structure |
| no_placeholder_text |
5 |
No placeholder text |
| technical_details_collapsed |
5 |
Tech details in |
LLM-Powered Transforms
Transform release notes for different audiences using Claude API:
# Requires ANTHROPIC_API_KEY environment variable
export ANTHROPIC_API_KEY="your-api-key"
# Create executive summary
node .claude/skills/release-notes/lib/transform-llm.cjs docs/release-notes/v1.1.0.md --transform executive
# Transform for business stakeholders
node .claude/skills/release-notes/lib/transform-llm.cjs docs/release-notes/v1.1.0.md --transform business --output docs/release-notes/v1.1.0-business.md
# Transform for end users
node .claude/skills/release-notes/lib/transform-llm.cjs docs/release-notes/v1.1.0.md --transform enduser
Transform Types:
| Type |
Description |
summarize |
Brief 3-5 bullet point summary |
business |
ROI-focused, business language |
enduser |
User-friendly, non-technical |
executive |
Strategic impact summary |
technical |
Enhanced technical details |
Full Enhanced Pipeline
Combine all features for comprehensive release notes:
# Enhanced pipeline with service detection
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD --with-files | \
node .claude/skills/release-notes/lib/detect-services.cjs | \
node .claude/skills/release-notes/lib/categorize-commits.cjs | \
node .claude/skills/release-notes/lib/detect-breaking.cjs | \
node .claude/skills/release-notes/lib/contributor-stats.cjs | \
node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0
# With version bumping and validation
node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD --with-files | \
node .claude/skills/release-notes/lib/bump-version.cjs | \
node .claude/skills/release-notes/lib/categorize-commits.cjs | \
node .claude/skills/release-notes/lib/render-template.cjs --output docs/release-notes/v1.1.0.md && \
node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md
Configuration
See config.yaml for:
- categories - Commit type to section mapping
- services - Service boundary detection by file patterns
- exclude - Patterns to exclude from user-facing notes
- output - Directory and filename format settings
Output Structure
# Release Notes: v1.1.0
**Date:** 2025-01-11
**Version:** v1.1.0
**Status:** Draft
---
## Summary
This release includes 3 new features, 2 improvements, 5 bug fixes.
## What's New
- **Add order export endpoint** (API)
- **Implement dark mode toggle** (UI)
## Improvements
- **Optimize database queries** (Persistence)
## Bug Fixes
- **Fix date picker timezone issue** (Frontend)
- **Resolve null pointer in auth flow**
## Documentation
- **Update API documentation** (API)
## Breaking Changes
> **Warning**: The following changes may require migration
### Migrate to OAuth 2.1 (Auth)
Legacy JWT tokens no longer accepted.
Migration guide: docs/migrations/oauth-2.1.md
---
## Technical Details
<details>
<summary>For Developers</summary>
### Commits Included
| Hash | Type | Description |
| ------- | ---- | ------------------------------ |
| abc1234 | feat | Add order export endpoint |
| def5678 | fix | Fix date picker timezone issue |
...
</details>
## Contributors
- @john.doe
- @jane.smith
---
_Generated by AI_
Human Review Gate
Generated release notes are Draft status by default:
- Review - Check accuracy, add context where needed
- Enhance - Add migration steps, links, screenshots
- Approve - Change status to "Released"
- Publish - Commit and push
Integration with Other Skills
$commit - After generating notes, commit them/git-manager - Create PR for release notes review$docs-update - Update CHANGELOG.md with new release
Troubleshooting
No commits found
Verify the refs exist and have commits between them:
git log --oneline <base>..<head>
Non-conventional commits
Commits not following type(scope): description format go to "other" category. Consider running commitlint enforcement.
Missing scope context
Add scope mappings to config.yaml → services section for better context labels.
[IMPORTANT] Use task tracking to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ATTENTION ask user whether to skip.
AI Mistake Prevention — Failure modes to avoid on every task:
Re-read files after context changes. Context compaction, resume, or long-running work can make memory stale; verify current files before acting.
Verify generated content against source evidence. AI hallucinates APIs, names, claims, and document facts. Check the relevant source before documenting or referencing.
Check downstream references before deleting or renaming. Removing an artifact can stale docs, generated mirrors, configs, and callers; map references first.
Trace the full impact chain after edits. Changing a definition can miss derived outputs and consumers. Follow the affected chain before declaring done.
Verify ALL affected outputs, not just the first. One green check is not all green checks; validate every output surface the change can affect.
Assume existing values are intentional — ask WHY before changing OR flagging one as a defect. Before changing or reporting a constant, limit, flag, cutoff, wording, or pattern, read nearby context and history, the CALLER's ordering, and 2+ sibling call sites of the same convention. A doc stating WHAT without WHY is missing rationale, not proof of a missing guard.
Surface ambiguity before acting — don't pick silently. Multiple valid interpretations require an explicit question or stated assumption with risk.
Assert the outcome your system owns, not the intermediate state your infrastructure owns. When verifying async work, assert the final business state — never the delivery/retry bookkeeping held in shared infrastructure that any co-running process can write. Such a check passes when run alone and flakes the moment anything else shares that infrastructure.
Keep shared guidance role-relevant. Universal guidance must help every receiving skill or agent; code-specific obligations belong only in code-specific protocols.
Critical Thinking Mindset — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act.
Anti-hallucination: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.
MUST ATTENTION apply critical + sequential thinking — every claim needs appropriate traced evidence (file:line for repo/code claims; source URL or artifact section for research, product, content, and docs claims); confidence >80% to act, <60% DO NOT recommend. Anti-hallucination: never present guess as fact, admit uncertainty freely, cross-reference independently, stay skeptical of own confidence.
MUST ATTENTION apply AI mistake prevention — verify generated content against evidence, trace downstream references before deleting or renaming, verify all affected outputs, re-read files after context loss, and surface ambiguity before acting.
Project Protocol Overlay — Before executing this skill, resolve any PROJECT overlay rules layered onto it: match this skill's name against the Target column of the project's skill-protocol index (docs/project-reference/skill-protocols-reference.md by default; a referenceDocs entry in docs/project-config.json overrides the path), taking the most specific matching tier ONLY — exact name > glob > *. That precedence orders overlays against EACH OTHER, never against this skill. Read ONLY the matched bodies, resolved as <protocols-dir>/<Name>.md; a row's Body link is display text, never a read path. A matched body that is missing or malformed is REPORTED and skipped — never reconstructed from the index Description. No index, or no match -> proceed with no overlay, silently. Full contract: .claude/skills/project-skill-protocol/references/registry.md.
Overlays are ADDITIVE ONLY: they ADD rules on top of this skill's own protocol and NEVER replace, override, disable, or reinterpret a rule it already states — removing every overlay must return this skill to exactly its documented behavior. An overlay is a BRIEF, not an authority escalation: it can NEVER waive a workflow gate, git discipline, a review gate, or a user-confirmation gate. A genuine overlay-vs-skill conflict, or two equally-specific overlays that directly contradict -> surface both to the user; NEVER resolve silently.
MUST ATTENTION resolve project protocol overlays for this skill BEFORE executing — most specific matching tier only (exact > glob > *, which ranks overlays against each other, NEVER against this skill), read only matched bodies at <protocols-dir>/<Name>.md; a missing or malformed body is reported, never reconstructed. Overlays are ADDITIVE ONLY (they never replace this skill's own rules) and are a brief, NEVER an authority escalation; an equal-specificity contradiction goes to the user.
Closing Reminders
Protocols in force (concise digest of the SYNC/shared blocks this skill carries):
- AI Mistake Prevention: verify generated content against evidence, trace downstream references, verify all affected outputs, re-read after context loss, surface ambiguity.
- Critical Thinking: apply critical + sequential thinking; every claim needs traced proof, confidence >80% to act.
IMPORTANT MUST ATTENTION break work into small todo tasks using task tracking BEFORE starting
IMPORTANT MUST ATTENTION search codebase for 3+ similar patterns before creating new code
IMPORTANT MUST ATTENTION cite file:line evidence for every claim (confidence >80% to act)
IMPORTANT MUST ATTENTION add a final review todo task to verify work quality
[TASK-PLANNING] Before acting, analyze task scope and systematically break it into small todo tasks and sub-tasks using task tracking.
Hookless Prompt Protocol Mirror (Auto-Synced)
Source: .claude/.ck.json + .claude/skills/shared/sync-inline-versions.md (:full blocks) + .claude/scripts/lib/hookless-prompt-protocol.cjs
[WORKFLOW-EXECUTION-PROTOCOL] [BLOCKING] Workflow Execution Protocol — MANDATORY IMPORTANT MUST CRITICAL. Do not skip for any reason.
Generic portability boundary: Reusable skills and protocol text stay project-neutral; project-specific conventions are discovered from docs/project-config.json and docs/project-reference/. Apply shared AI-SDD from shared/sdd-artifact-contract.md. Read docs/project-config.json and docs/project-reference/docs-index-reference.md, then open the project reference docs named there. For spec, test-case, behavior-change, public-contract, or docs/specs/ work, route through the local spec docs named by the docs index: feature-spec-reference.md, spec-system-reference.md, spec-principles.md, and workflow-spec-test-code-cycle-reference.md when specs/tests/code must stay synchronized. If either file or a required reference doc is missing or stale, auto-run $project-init (or the narrow lower-level route such as $project-config, $docs-init, $scan-all, or $scan --target=<key>) before ordinary project-specific work. Any supported AI tool may execute when this shared context and local docs are available.
- DETECT: If the prompt starts with an explicit slash skill/workflow command, execute it directly. Otherwise match the prompt against the workflow catalog and skill list.
- ANALYZE: Choose the best option: execute directly, invoke a skill, activate a standard workflow, or compose a custom step combination.
- AUTO-SELECT: Pick the best option yourself. Do not ask the user to choose between direct execution, skill, standard workflow, or custom workflow.
- ACTIVATE: For a selected workflow, call
$start-workflow <workflowId>; for a selected skill, invoke that skill; for a custom workflow, sequence custom steps directly; for direct execution, proceed with the task.
- CREATE TASKS: task tracking for ALL workflow/skill/custom steps before execution when the selected path has multiple steps.
- PARALLELIZE: Before executing the task list, tag each task
PAR (independent inputs + write set disjoint from every other PAR task) or SEQ (name the blocking dependency), group PAR tasks into waves, declare the wave plan, and spawn each wave's sub-agents in ONE message — all-return barrier per wave, fan-out one level deep unless a sub-agent's own definition authorizes further fan-out. Sequential-by-default is a defect when tasks are independent; do not parallelize shared write targets, output-consuming tasks, trivial single-file work, ordering a skill or workflow explicitly fixes, or user-approval gates.
- EXECUTE: Advance per the Workflow Step Advancement & Parallel Phases rule in your context instructions — model-driven; a sub-agent completion advances a step identically to an inline call; a parallel-phase group is an all-return barrier (advance only after ALL members return, never serialize it)
Shared AI-SDD Protocol Markers
Source: .claude/skills/shared/sync-inline-versions.md
SYNC:ai-sdd-artifact-contract
AI-SDD Artifact Contract — Shared spec-driven development rules stay portable and source-owned.
- Keep reusable AI-SDD principles in
.claude; put repository-specific paths, commands, owners, products, and formats in project config/reference docs.
- Preserve cycle:
spec -> plan -> tasks -> implement -> verify -> update spec/docs.
- Trace every requirement or invariant through decision, task, TC/test, source evidence, and docs/spec update.
- Treat code-to-spec extraction as reference-only until accepted by the canonical spec owner.
- Any supported AI tool may plan, implement, review, or verify with synced context; using multiple tools is optional.
- Update
.claude source first, then sync generated mirrors; do not manually edit .agents, .codex, or AGENTS.md. — why: mirrors are generated artifacts; hand-edits are overwritten on the next sync
- If
docs/project-config.json, root instruction files, or a required project-reference doc is missing or stale, auto-run $project-init or the narrow lower-level route before ordinary project-specific work.
Active reference: shared/sdd-artifact-contract.md in the active skills root.
SYNC:ai-sdd-artifact-contract:reminder
- MANDATORY Apply
shared/sdd-artifact-contract.md; keep reusable AI-SDD in .claude and local rules in project docs.
- MANDATORY Code-to-spec extraction is reference-only until canonical acceptance; any supported AI tool may execute with synced context.
- MANDATORY Update
.claude source before syncing generated mirrors; do not manually edit .agents, .codex, or AGENTS.md.
- MANDATORY Missing or stale project config, root instruction files, or required reference docs route project-specific work through
$project-init or the narrow setup route automatically.
[TASK-PLANNING] [MANDATORY] BEFORE executing any workflow or skill step, create/update task tracking for all planned steps, then keep it synchronized as each step starts/completes.
[LESSON-LEARNED-REMINDER] [BLOCKING] Task Planning & Continuous Improvement — MANDATORY. Do not skip.
Break work into small tasks (task tracking) before starting. Add final task: "Analyze AI mistakes & lessons learned".
Extract lessons — ROOT CAUSE ONLY, not symptom fixes:
- Name the FAILURE MODE (reasoning/assumption failure), not symptom — "assumed API existed without reading source" not "used wrong enum value".
- Generality test: does this failure mode apply to ≥3 contexts/codebases? If not, abstract one level up.
- Write as a universal rule — strip project-specific names/paths/classes. Useful on any codebase.
- Consolidate: multiple mistakes sharing one failure mode → ONE lesson.
- Recurrence gate: "Would this recur in future session WITHOUT this reminder?" — No → skip
$learn.
- Auto-fix gate: "Could
$code-review/$code-simplifier/$security-review/$lint catch this?" — Yes → improve review skill instead.
- BOTH gates pass → ask user to run
$learn.
[CRITICAL-THINKING-MINDSET] Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act.
Anti-hallucination principle: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.
AI Attention principle (Primacy-Recency): Put the 3 most critical rules at both top and bottom of long prompts/protocols so instruction adherence survives long context windows.
Goal-driven execution: Define success criteria first, loop until verified, and stop only when observable checks pass.
Tests verify intent: Tests must protect business rules/invariants and fail when the protected intent breaks, not only mirror current behavior.
Common AI Mistake Prevention (System Lessons)
- Re-read files after context compaction. Edit requires prior Read in same context; compaction wipes read state. Re-read before editing.
- Grep for old terms after bulk replacements. AI over-trusts find/replace completeness. Grep full repo after bulk edits for missed refs in docs/configs/catalogs.
- Check downstream references before deleting. Deletions cascade doc/code staleness. Map referencing files before removal.
- After memory loss, check existing state before creating new. Compaction wipes prior-work memory. Query current state to resume — never blindly duplicate.
- Verify AI-generated content against actual code. AI hallucinates APIs, class names, method signatures. Grep to confirm existence before documenting/referencing.
- Trace full dependency chain after edits. Changing a definition misses downstream consumers. Trace the full chain.
- When renaming, grep ALL consumer file types. Some file types silently ignore missing refs (no compile error). Search code, templates, configs, generated files.
- Trace ALL code paths when verifying correctness. Code existing ≠ code executing. Trace early exits, error branches, conditional skips — not just happy path.
- Update docs that embed canonical data when source changes. Docs inlining derived data (workflows, schemas, configs) go stale silently. Update all embedding docs alongside source.
- Verify sub-agent results after context recovery. Background agents may finish while parent compacted — grep-verify output, don't trust assumed completion.
- Cross-check full target list against sub-agent assignments. Parallel sub-agents by category miss boundary items. Reconcile union of assignments against target list before proceeding.
- Sub-agents inherit knowledge only from their agent .md definition — use custom agent types, not built-in Explore. Tool adoption = permission + knowledge + enforcement (numbered workflow step).
- Persist sub-agent findings incrementally, not as a final batch. Long sub-agents hit cutoffs before final write — findings lost. Instruct append-per-section to report file.
- When debugging, ask "whose responsibility?" before fixing. Trace caller (wrong data) vs callee (wrong handling). Fix at responsible layer — never patch symptom site.
- Test failure → record a provisional verdict before trace/edit, then investigate. Use the full five-way taxonomy: SOURCE-WRONG (production violates intent), TEST-WRONG (assertion/setup is stale), TEST-NOT-OPTIMAL (valid but fragile or low-signal test), ENVIRONMENT-BLOCKED (external state prevents a verdict), or AMBIGUOUS (intent/evidence cannot choose safely). Then trace root cause and triangulate against the governing spec (
docs/specs/** if one exists) AND source. NEVER weaken an assertion, add a skip, relax a timeout, or change source merely to force green.
- Grep ALL removed names after extraction/refactoring. Primary file "done" ≠ secondary files clean. Grep entire scope for every removed symbol before declaring complete.
- Assume existing values are intentional — ask WHY before changing OR flagging one as a defect. Pattern-matching as "wrong" skips context. Before changing or reporting any constant/limit/flag/cutoff: read comments, git blame, the CALLER's ordering (the guarantee that makes the value correct usually lives in code running immediately BEFORE the cited line), and 2+ sibling call sites of the same convention. A doc stating WHAT without WHY is missing rationale, not proof of a missing guard — and in a validation pass, an accurate
file:line citation proves the transcription, never the defect.
- Verify ALL affected outputs, not just the first. One build green ≠ all green. Multi-stack changes (backend/frontend/tests/docs) require verifying EVERY output.
- Evaluate fit before copying a nearby pattern. Closest example ≠ matching preconditions — verify the new context shares the same constraints, base classes, scope, lifetime.
- Holistic-first debugging — resist nearest-attention trap. Don't dive into first plausible cause. List EVERY precondition (config, env vars, paths, DB, endpoints, creds, versions, DI, data). Verify each against evidence (grep/query — not reasoning). Ask "what would falsify this?" — if nothing, it's not a hypothesis. Most expensive failure: going deeper in "obvious" layer while bug sits in layer never questioned.
- Surgical changes — apply the diff test (context-aware). Two modes: (1) Bug fix → every line traces to the bug; no restyling; orphan cleanup only for imports YOUR changes made unused. (2) Review/enhancement → implement improvements AND announce as "Enhancement beyond main request: [what]". Never silently scope-creep. Diff test: "Would this line exist if I wasn't asked to do X?" — if no, delete or announce.
- Surface ambiguity before coding — don't pick silently. Multiple valid interpretations → present each with effort: "[Request] could mean (1) [N h], (2) [N h]. Which matters?" List scope/format/volume/constraints assumptions first. If simpler path exists, say so. Never silently pick.
- [MANDATORY FIRST ACTION] ALWAYS activate a suitable skill or workflow BEFORE responding. Match task against workflow catalog + skill list; invoke via skill invocation or
$start-workflow <workflowId>. NEVER answer or write code before checking. Skip = protocol violation.
- Why-Review adversarial mindset — apply when reviewing any plan, decision, or design. Default SKEPTIC not VALIDATOR: steel-man a rejected alternative, invert each stated reason ("what does it sacrifice?"), stress-test top 2-3 assumptions, run pre-mortem ("ships, fails in 3 months — what breaks?"), surface 1-2 alternatives author missed. Section presence ≠ quality; quality = causal reasoning + concrete mitigations + evidence, not "it's better" or "monitor closely".
- Front-load report-write in sub-agent prompts for large reviews. Many-file sub-agents hit budget before final write — findings lost. Design prompts so: (1) report-write is first explicit deliverable, (2) append per-file/section (not batched), (3) scope bounded so reads don't exhaust budget. Truncated mid-sentence with no report file → spawn narrower scope, don't retry same prompt.
- After context compaction, re-verify all prior phase outcomes before continuing. Summaries describe intent, not environment state (git index, filesystem, processes). On resume, FIRST audit: git status, re-read modified files, verify filesystem. Every "completed" claim is an untested hypothesis until evidence confirms.
- OOM/memory: check row count before row size. Triage: (1) Unbounded query — no DB filter for trigger? Push filter to DB; eliminates OOM. (2) Large rows? Projection reduces proportionally. Row reduction > projection in ROI.
- Assert the outcome your system OWNS, never the intermediate state your INFRASTRUCTURE owns. When testing anything asynchronous (queue/broker delivery, retries, background jobs, caches, replication), assert the final business/entity state. NEVER assert the delivery bookkeeping — consume/send status, attempt counts, last-error, row existence or counts in a broker, scheduler, or outbox/inbox table. That bookkeeping lives in shared infrastructure that ANY co-running process (a peer worker, a second replica, a leftover local container) can write, usually under a deterministic shared key, so the assertion silently tests the developer's environment instead of the system: green when run alone, flaky the instant anything else shares that broker + database. Gate question for every assertion: "would this hold no matter WHICH process did the work?" — if no, assert the converged data state instead. Corollary: process-local fault injection and in-process telemetry cannot gate work any process may perform — use them as stress amplifiers (arm → bounded window → disarm → assert convergence), never as preconditions.
- Keep domain concepts out of generic/shared/infrastructure layers. Reusable layer (shared library, framework, infra module) must reference NO consumer-specific domain concept — tenant/customer/product IDs, business entities, feature rules. Leak compiles + runs → passes review silently while coupling the "reusable" layer to one consumer. Keep shared type domain-free; push domain fields/logic down into the consumer via subclass/composition. — why: a layer coupled to one consumer's domain is no longer reusable.
1---2name: release-notes3description: [Git] Use when creating release notes from git history.4---5
6> Codex compatibility note:
7>
8> - Invoke repository skills with `$skill-name` in Codex; this mirrored copy rewrites legacy Claude `/skill-name` references.
9> - Task tracker mandate: BEFORE executing any workflow or skill step, create/update task tracking for all steps and keep it synchronized as progress changes.
10> - User-question prompts mean to ask the user directly in Codex.
11> - Ignore Claude-specific mode-switch instructions when they appear.
12> - Strict execution contract: when a user explicitly invokes a skill, execute that skill protocol as written.
13> - Subagent authorization: when a skill is user-invoked or AI-detected and its protocol requires subagents, that skill activation authorizes use of the required `spawn_agent` subagent(s) for that task.
14> - Do not skip, reorder, or merge protocol steps unless the user explicitly approves the deviation first.
15> - For workflow skills, execute each listed child-skill step explicitly and report step-by-step evidence.
16> - If a required step/tool cannot run in this environment, stop and ask the user before adapting.
17
18<!-- CODEX:PROJECT-REFERENCE-LOADING:START -->
19
20## Codex Project-Reference Loading (No Hooks)
21
22Codex uses static project-reference loading instead of runtime-injected project docs.
23When coding, planning, debugging, testing, or reviewing, open project docs explicitly using this routing.
24
25**Always read:**
26
27- `docs/project-config.json` (project-specific paths, commands, modules, and workflow/test settings)
28- `docs/project-reference/docs-index-reference.md` (routes to the full `docs/project-reference/*` catalog)
29- `docs/project-reference/lessons.md` (always-on guardrails and anti-patterns)
30
31**Missing/stale context route:** If `docs/project-config.json`, the docs index, `lessons.md`, `CLAUDE.md`, `AGENTS.md`, or any task-required reference doc is missing or stale, auto-run `$project-init` or the narrow setup route (`$project-config`, `$docs-init`, `$scan-all`, `$scan --target=<key>`, `$claude-md-init`) before ordinary project-specific work. If Codex mirrors or `AGENTS.md` are missing/stale, ask the user to run `$sync-codex`; do not auto-run it.
32
33**Situation-based docs:**
34
35- Project structure/architecture/tech-stack/deployment/setup (any layer — backend, frontend, or infra): `project-structure-reference.md`
36- Backend/CQRS/API/domain/entity changes: `backend-patterns-reference.md`, `domain-entities-reference.md`
37- Frontend/UI/styling/design-system: `frontend-patterns-reference.md`, `scss-styling-guide.md`, `design-system/README.md`
38- Spec authoring, `docs/specs/` pathing, or TC format: `feature-spec-reference.md`, `spec-system-reference.md`, `spec-principles.md`
39- Behavior/public-contract changes or spec-test-code sync: `workflow-spec-test-code-cycle-reference.md` plus the spec docs above
40- Derived spec indexes/ERDs/reimplementation guides: `spec-system-reference.md` and source Feature Specs under `docs/specs/`
41- Integration test implementation/review: `integration-test-reference.md`
42- E2E test implementation/review: `e2e-test-reference.md`
43- Code review/audit work: `code-review-rules.md` plus domain docs above based on changed files
44
45Do not read all docs blindly. Start from `docs-index-reference.md`, then open only relevant files for the task.
46
47<!-- CODEX:PROJECT-REFERENCE-LOADING:END -->
48
49## Quick Summary
50
51**Goal:** Generate professional release notes from git commits with automated categorization, service detection, and validation.
52
53**Workflow:**
54
551. **Parse Commits** — `parse-commits.cjs <base> <head>` extracts structured data from git
562. **Categorize** — Pipe through `categorize-commits.cjs` for user-facing vs internal sections
573. **Render** — `render-template.cjs --version vX.Y.Z` generates markdown with Summary, What's New, Improvements, Bug Fixes, Breaking Changes, Technical Details
58
59**Key Rules:**
60
61- **Pipeline**: parse → categorize → render → validate → transform
62- **Advanced**: Service detection, breaking change analysis, PR metadata, contributor stats, version bumping
63- **Human Review**: Generated notes are Draft status, require review/enhance/approve before publish
64- **Validation**: `validate-notes.cjs` scores against quality rules (100 points)
65
66**Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).**
67
68# Release Notes Generation Skill
69
70Generate professional release notes from git commits between two refs with automated categorization.
71
72## Invocation
73
74```
75$release-notes [base] [head] [--version vX.Y.Z] [--output path]
76```
77
78**Examples:**
79
80```bash
81# Generate release notes for commits since last tag
82$release-notes v1.0.0 HEAD --version v1.1.0
83
84# Compare branches
85$release-notes main feature/new-auth --version v2.0.0-beta
86
87# Output to specific file
88$release-notes v1.0.0 HEAD --version v1.1.0 --output docs/release-notes/250111-v1.1.0.md
89```
90
91## Workflow
92
93### Step 1: Parse Commits
94
95Execute the commit parser to extract structured data from git history:
96
97```bash
98node .claude/skills/release-notes/lib/parse-commits.cjs <base> <head> [--with-files]
99```
100
101**Output:** JSON with commits array containing:
102
103- `hash`, `shortHash` - Commit identifiers
104- `type`, `scope`, `description` - Conventional commit parts
105- `breaking` - Boolean for breaking changes
106- `author`, `date` - Attribution
107- `files` - Changed files (with `--with-files` flag)
108
109### Step 2: Categorize Commits
110
111Pipe parsed commits through the categorizer:
112
113```bash
114node .claude/skills/release-notes/lib/parse-commits.cjs <base> <head> | \
115node .claude/skills/release-notes/lib/categorize-commits.cjs
116```
117
118**Categorization Rules:**
119| Type | Category | User-Facing |
120| --------------------------------------- | ------------ | --------------------- |
121| `feat` | features | Yes |
122| `fix` | fixes | Yes |
123| `perf` | improvements | Yes |
124| `docs` | docs | Yes (unless internal) |
125| `refactor` | improvements | Technical only |
126| `test`, `ci`, `build`, `chore`, `style` | internal | No |
127
128**Excluded Patterns:**
129
130- `chore(deps):` - Dependency updates
131- `chore(config):` - Configuration changes
132- `[skip changelog]` - Explicit skip
133- `[ci skip]` - CI markers
134
135### Step 3: Render Markdown
136
137Generate the final release notes document:
138
139```bash
140node .claude/skills/release-notes/lib/parse-commits.cjs <base> <head> | \
141node .claude/skills/release-notes/lib/categorize-commits.cjs | \
142node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0 --output docs/release-notes/250111-v1.1.0.md
143```
144
145## Complete Pipeline
146
147For generating release notes in a single command:
148
149```bash
150# Full pipeline with output to file
151node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
152node .claude/skills/release-notes/lib/categorize-commits.cjs | \
153node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0 --output docs/release-notes/250111-v1.1.0.md
154
155# Pipeline to stdout for review
156node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
157node .claude/skills/release-notes/lib/categorize-commits.cjs | \
158node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0
159```
160
161## Advanced Features
162
163### Service Boundary Detection
164
165Analyze which services are affected by the release:
166
167```bash
168# Parse with file changes, then detect services
169node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD --with-files | \
170node .claude/skills/release-notes/lib/detect-services.cjs
171```
172
173**Output:** Service impact analysis with severity levels (critical, high, medium, low)
174
175### Breaking Change Analysis
176
177Enhanced breaking change detection with migration info extraction:
178
179```bash
180node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
181node .claude/skills/release-notes/lib/categorize-commits.cjs | \
182node .claude/skills/release-notes/lib/detect-breaking.cjs
183```
184
185**Detects:**
186
187- `BREAKING CHANGE:` in commit body
188- `!` suffix on commit type (e.g., `feat!:`)
189- Migration instructions
190
191### PR Metadata Extraction
192
193Extract and link pull request information:
194
195```bash
196# Extract PR numbers from commit messages
197node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
198node .claude/skills/release-notes/lib/extract-pr-metadata.cjs
199
200# With GitHub API enrichment (requires gh CLI)
201node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
202node .claude/skills/release-notes/lib/extract-pr-metadata.cjs --fetch-gh
203```
204
205**Extracts:** PR numbers, titles, labels, authors from commits
206
207### Contributor Statistics
208
209Generate detailed contributor stats:
210
211```bash
212node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
213node .claude/skills/release-notes/lib/contributor-stats.cjs
214```
215
216**Output:** Contributor list with commit counts, feature/fix breakdown
217
218### Version Bumping
219
220Automatically determine and bump semantic version based on commit types:
221
222```bash
223# Auto-bump based on commits (feat→minor, fix→patch, BREAKING→major)
224node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD | \
225node .claude/skills/release-notes/lib/bump-version.cjs
226
227# Bump with prerelease tag
228node .claude/skills/release-notes/lib/bump-version.cjs --prerelease beta
229
230# Per-service versioning
231node .claude/skills/release-notes/lib/bump-version.cjs --service {service-name}
232
233# Dry run (don't write version file)
234node .claude/skills/release-notes/lib/bump-version.cjs --dry-run
235```
236
237**Version Files:**
238
239- Root: `.version`
240- Per-service: `.versions/<service-name>.version`
241
242### Quality Validation
243
244Validate release notes against quality rules:
245
246```bash
247# Validate with default threshold (70)
248node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md
249
250# Custom threshold
251node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md --threshold 80
252
253# JSON output for CI
254node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md --json
255```
256
257**Validation Rules (100 points total):**
258| Rule | Weight | Description |
259| --------------------------- | ------ | ---------------------------- |
260| summary_exists | 15 | Has Summary section |
261| summary_not_empty | 10 | Summary has content |
262| has_version | 10 | Version number present |
263| features_documented | 10 | Features properly formatted |
264| fixes_documented | 10 | Bug fixes properly formatted |
265| no_broken_links | 10 | No empty link references |
266| contributors_listed | 10 | Contributors section present |
267| has_date | 5 | Date present |
268| no_todo_markers | 5 | No TODO/FIXME markers |
269| proper_heading_hierarchy | 5 | Proper H1→H2 structure |
270| no_placeholder_text | 5 | No placeholder text |
271| technical_details_collapsed | 5 | Tech details in <details> |
272
273### LLM-Powered Transforms
274
275Transform release notes for different audiences using Claude API:
276
277```bash
278# Requires ANTHROPIC_API_KEY environment variable
279export ANTHROPIC_API_KEY="your-api-key"
280
281# Create executive summary
282node .claude/skills/release-notes/lib/transform-llm.cjs docs/release-notes/v1.1.0.md --transform executive
283
284# Transform for business stakeholders
285node .claude/skills/release-notes/lib/transform-llm.cjs docs/release-notes/v1.1.0.md --transform business --output docs/release-notes/v1.1.0-business.md
286
287# Transform for end users
288node .claude/skills/release-notes/lib/transform-llm.cjs docs/release-notes/v1.1.0.md --transform enduser
289```
290
291**Transform Types:**
292| Type | Description |
293| ----------- | ------------------------------ |
294| `summarize` | Brief 3-5 bullet point summary |
295| `business` | ROI-focused, business language |
296| `enduser` | User-friendly, non-technical |
297| `executive` | Strategic impact summary |
298| `technical` | Enhanced technical details |
299
300### Full Enhanced Pipeline
301
302Combine all features for comprehensive release notes:
303
304```bash
305# Enhanced pipeline with service detection
306node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD --with-files | \
307node .claude/skills/release-notes/lib/detect-services.cjs | \
308node .claude/skills/release-notes/lib/categorize-commits.cjs | \
309node .claude/skills/release-notes/lib/detect-breaking.cjs | \
310node .claude/skills/release-notes/lib/contributor-stats.cjs | \
311node .claude/skills/release-notes/lib/render-template.cjs --version v1.1.0
312
313# With version bumping and validation
314node .claude/skills/release-notes/lib/parse-commits.cjs v1.0.0 HEAD --with-files | \
315node .claude/skills/release-notes/lib/bump-version.cjs | \
316node .claude/skills/release-notes/lib/categorize-commits.cjs | \
317node .claude/skills/release-notes/lib/render-template.cjs --output docs/release-notes/v1.1.0.md && \
318node .claude/skills/release-notes/lib/validate-notes.cjs docs/release-notes/v1.1.0.md
319```
320
321## Configuration
322
323See `config.yaml` for:
324
325- **categories** - Commit type to section mapping
326- **services** - Service boundary detection by file patterns
327- **exclude** - Patterns to exclude from user-facing notes
328- **output** - Directory and filename format settings
329
330## Output Structure
331
332```markdown
333# Release Notes: v1.1.0
334
335**Date:** 2025-01-11
336**Version:** v1.1.0
337**Status:** Draft
338
339---
340
341## Summary
342
343This release includes 3 new features, 2 improvements, 5 bug fixes.
344
345## What's New
346
347- **Add order export endpoint** (API)
348- **Implement dark mode toggle** (UI)
349
350## Improvements
351
352- **Optimize database queries** (Persistence)
353
354## Bug Fixes
355
356- **Fix date picker timezone issue** (Frontend)
357- **Resolve null pointer in auth flow**
358
359## Documentation
360
361- **Update API documentation** (API)
362
363## Breaking Changes
364
365> **Warning**: The following changes may require migration
366
367### Migrate to OAuth 2.1 (Auth)
368
369Legacy JWT tokens no longer accepted.
370Migration guide: docs/migrations/oauth-2.1.md
371
372---
373
374## Technical Details
375
376<details>
377<summary>For Developers</summary>
378
379### Commits Included
380
381| Hash | Type | Description |
382| ------- | ---- | ------------------------------ |
383| abc1234 | feat | Add order export endpoint |
384| def5678 | fix | Fix date picker timezone issue |
385
386...
387
388</details>
389
390## Contributors
391
392- @john.doe
393- @jane.smith
394
395---
396
397_Generated by AI_
398```
399
400## Human Review Gate
401
402Generated release notes are **Draft** status by default:
403
4041. **Review** - Check accuracy, add context where needed
4052. **Enhance** - Add migration steps, links, screenshots
4063. **Approve** - Change status to "Released"
4074. **Publish** - Commit and push
408
409## Integration with Other Skills
410
411- **`$commit`** - After generating notes, commit them
412- **`/git-manager`** - Create PR for release notes review
413- **`$docs-update`** - Update CHANGELOG.md with new release
414
415## Troubleshooting
416
417### No commits found
418
419Verify the refs exist and have commits between them:
420
421```bash
422git log --oneline <base>..<head>
423```
424
425### Non-conventional commits
426
427Commits not following `type(scope): description` format go to "other" category. Consider running commitlint enforcement.
428
429### Missing scope context
430
431Add scope mappings to `config.yaml` → `services` section for better context labels.
432
433---
434
435> **[IMPORTANT]** Use task tracking to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ATTENTION ask user whether to skip.
436
437<!-- SYNC:ai-mistake-prevention -->
438
439> **AI Mistake Prevention** — Failure modes to avoid on every task:
440>
441> **Re-read files after context changes.** Context compaction, resume, or long-running work can make memory stale; verify current files before acting.
442> **Verify generated content against source evidence.** AI hallucinates APIs, names, claims, and document facts. Check the relevant source before documenting or referencing.
443> **Check downstream references before deleting or renaming.** Removing an artifact can stale docs, generated mirrors, configs, and callers; map references first.
444> **Trace the full impact chain after edits.** Changing a definition can miss derived outputs and consumers. Follow the affected chain before declaring done.
445> **Verify ALL affected outputs, not just the first.** One green check is not all green checks; validate every output surface the change can affect.
446> **Assume existing values are intentional — ask WHY before changing OR flagging one as a defect.** Before changing or reporting a constant, limit, flag, cutoff, wording, or pattern, read nearby context and history, the CALLER's ordering, and 2+ sibling call sites of the same convention. A doc stating WHAT without WHY is missing rationale, not proof of a missing guard.
447> **Surface ambiguity before acting — don't pick silently.** Multiple valid interpretations require an explicit question or stated assumption with risk.
448> **Assert the outcome your system owns, not the intermediate state your infrastructure owns.** When verifying async work, assert the final business state — never the delivery/retry bookkeeping held in shared infrastructure that any co-running process can write. Such a check passes when run alone and flakes the moment anything else shares that infrastructure.
449> **Keep shared guidance role-relevant.** Universal guidance must help every receiving skill or agent; code-specific obligations belong only in code-specific protocols.
450
451<!-- /SYNC:ai-mistake-prevention -->
452
453<!-- SYNC:critical-thinking-mindset -->
454
455> **Critical Thinking Mindset** — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act.
456> **Anti-hallucination:** Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.
457
458<!-- /SYNC:critical-thinking-mindset -->
459
460<!-- SYNC:critical-thinking-mindset:reminder -->
461
462**MUST ATTENTION** apply critical + sequential thinking — every claim needs appropriate traced evidence (`file:line` for repo/code claims; source URL or artifact section for research, product, content, and docs claims); confidence >80% to act, <60% DO NOT recommend. Anti-hallucination: never present guess as fact, admit uncertainty freely, cross-reference independently, stay skeptical of own confidence.
463
464<!-- /SYNC:critical-thinking-mindset:reminder -->
465
466<!-- SYNC:ai-mistake-prevention:reminder -->
467
468**MUST ATTENTION** apply AI mistake prevention — verify generated content against evidence, trace downstream references before deleting or renaming, verify all affected outputs, re-read files after context loss, and surface ambiguity before acting.
469
470<!-- /SYNC:ai-mistake-prevention:reminder -->
471
472<!-- SYNC:project-protocol-overlay -->
473
474> **Project Protocol Overlay** — Before executing this skill, resolve any PROJECT overlay rules layered onto it: match this skill's name against the `Target` column of the project's skill-protocol index (`docs/project-reference/skill-protocols-reference.md` by default; a `referenceDocs` entry in `docs/project-config.json` overrides the path), taking the most specific matching tier ONLY — exact name > glob > `*`. **That precedence orders overlays against EACH OTHER, never against this skill.** Read ONLY the matched bodies, resolved as `<protocols-dir>/<Name>.md`; a row's Body link is display text, never a read path. A matched body that is missing or malformed is REPORTED and skipped — never reconstructed from the index Description. No index, or no match -> proceed with no overlay, silently. Full contract: `.claude/skills/project-skill-protocol/references/registry.md`.
475>
476> Overlays are **ADDITIVE ONLY**: they ADD rules on top of this skill's own protocol and NEVER replace, override, disable, or reinterpret a rule it already states — removing every overlay must return this skill to exactly its documented behavior. An overlay is a BRIEF, not an authority escalation: it can NEVER waive a workflow gate, git discipline, a review gate, or a user-confirmation gate. A genuine overlay-vs-skill conflict, or two equally-specific overlays that directly contradict -> surface both to the user; NEVER resolve silently.
477
478<!-- /SYNC:project-protocol-overlay -->
479
480<!-- SYNC:project-protocol-overlay:reminder -->
481
482**MUST ATTENTION** resolve project protocol overlays for this skill BEFORE executing — most specific matching tier only (exact > glob > `*`, which ranks overlays against each other, NEVER against this skill), read only matched bodies at `<protocols-dir>/<Name>.md`; a missing or malformed body is reported, never reconstructed. Overlays are ADDITIVE ONLY (they never replace this skill's own rules) and are a brief, NEVER an authority escalation; an equal-specificity contradiction goes to the user.
483
484<!-- /SYNC:project-protocol-overlay:reminder -->
485
486## Closing Reminders
487
488**Protocols in force (concise digest of the SYNC/shared blocks this skill carries):**
489
490- **AI Mistake Prevention:** verify generated content against evidence, trace downstream references, verify all affected outputs, re-read after context loss, surface ambiguity.
491- **Critical Thinking:** apply critical + sequential thinking; every claim needs traced proof, confidence >80% to act.
492
493**IMPORTANT MUST ATTENTION** break work into small todo tasks using task tracking BEFORE starting
494**IMPORTANT MUST ATTENTION** search codebase for 3+ similar patterns before creating new code
495**IMPORTANT MUST ATTENTION** cite `file:line` evidence for every claim (confidence >80% to act)
496**IMPORTANT MUST ATTENTION** add a final review todo task to verify work quality
497
498**[TASK-PLANNING]** Before acting, analyze task scope and systematically break it into small todo tasks and sub-tasks using task tracking.
499
500<!-- CODEX:SYNC-PROMPT-PROTOCOLS:START -->
501
502## Hookless Prompt Protocol Mirror (Auto-Synced)
503
504Source: `.claude/.ck.json` + `.claude/skills/shared/sync-inline-versions.md` (`:full` blocks) + `.claude/scripts/lib/hookless-prompt-protocol.cjs`
505
506## [WORKFLOW-EXECUTION-PROTOCOL] [BLOCKING] Workflow Execution Protocol — MANDATORY IMPORTANT MUST CRITICAL. Do not skip for any reason.
507
508**Generic portability boundary:** Reusable skills and protocol text stay project-neutral; project-specific conventions are discovered from docs/project-config.json and docs/project-reference/. Apply shared AI-SDD from `shared/sdd-artifact-contract.md`. Read `docs/project-config.json` and `docs/project-reference/docs-index-reference.md`, then open the project reference docs named there. For spec, test-case, behavior-change, public-contract, or `docs/specs/` work, route through the local spec docs named by the docs index: `feature-spec-reference.md`, `spec-system-reference.md`, `spec-principles.md`, and `workflow-spec-test-code-cycle-reference.md` when specs/tests/code must stay synchronized. If either file or a required reference doc is missing or stale, auto-run `$project-init` (or the narrow lower-level route such as `$project-config`, `$docs-init`, `$scan-all`, or `$scan --target=<key>`) before ordinary project-specific work. Any supported AI tool may execute when this shared context and local docs are available.
509
5101. **DETECT:** If the prompt starts with an explicit slash skill/workflow command, execute it directly. Otherwise match the prompt against the workflow catalog and skill list.
5112. **ANALYZE:** Choose the best option: execute directly, invoke a skill, activate a standard workflow, or compose a custom step combination.
5123. **AUTO-SELECT:** Pick the best option yourself. Do not ask the user to choose between direct execution, skill, standard workflow, or custom workflow.
5134. **ACTIVATE:** For a selected workflow, call `$start-workflow <workflowId>`; for a selected skill, invoke that skill; for a custom workflow, sequence custom steps directly; for direct execution, proceed with the task.
5145. **CREATE TASKS:** task tracking for ALL workflow/skill/custom steps before execution when the selected path has multiple steps.
5156. **PARALLELIZE:** Before executing the task list, tag each task `PAR` (independent inputs + write set disjoint from every other `PAR` task) or `SEQ` (name the blocking dependency), group `PAR` tasks into waves, declare the wave plan, and spawn each wave's sub-agents in ONE message — all-return barrier per wave, fan-out one level deep unless a sub-agent's own definition authorizes further fan-out. Sequential-by-default is a defect when tasks are independent; do not parallelize shared write targets, output-consuming tasks, trivial single-file work, ordering a skill or workflow explicitly fixes, or user-approval gates.
5167. **EXECUTE:** Advance per the **Workflow Step Advancement & Parallel Phases** rule in your context instructions — model-driven; a sub-agent completion advances a step identically to an inline call; a parallel-phase group is an all-return barrier (advance only after ALL members return, never serialize it)
517
518## Shared AI-SDD Protocol Markers
519
520Source: `.claude/skills/shared/sync-inline-versions.md`
521
522## SYNC:ai-sdd-artifact-contract
523
524> **AI-SDD Artifact Contract** — Shared spec-driven development rules stay portable and source-owned.
525>
526> 1. Keep reusable AI-SDD principles in `.claude`; put repository-specific paths, commands, owners, products, and formats in project config/reference docs.
527> 2. Preserve cycle: `spec -> plan -> tasks -> implement -> verify -> update spec/docs`.
528> 3. Trace every requirement or invariant through decision, task, TC/test, source evidence, and docs/spec update.
529> 4. Treat code-to-spec extraction as reference-only until accepted by the canonical spec owner.
530> 5. Any supported AI tool may plan, implement, review, or verify with synced context; using multiple tools is optional.
531> 6. Update `.claude` source first, then sync generated mirrors; do not manually edit `.agents`, `.codex`, or `AGENTS.md`. — why: mirrors are generated artifacts; hand-edits are overwritten on the next sync
532> 7. If `docs/project-config.json`, root instruction files, or a required project-reference doc is missing or stale, auto-run `$project-init` or the narrow lower-level route before ordinary project-specific work.
533>
534> **Active reference:** `shared/sdd-artifact-contract.md` in the active skills root.
535
536---
537
538## SYNC:ai-sdd-artifact-contract:reminder
539
540- **MANDATORY** Apply `shared/sdd-artifact-contract.md`; keep reusable AI-SDD in `.claude` and local rules in project docs.
541- **MANDATORY** Code-to-spec extraction is reference-only until canonical acceptance; any supported AI tool may execute with synced context.
542- **MANDATORY** Update `.claude` source before syncing generated mirrors; do not manually edit `.agents`, `.codex`, or `AGENTS.md`.
543- **MANDATORY** Missing or stale project config, root instruction files, or required reference docs route project-specific work through `$project-init` or the narrow setup route automatically.
544 **[TASK-PLANNING] [MANDATORY]** BEFORE executing any workflow or skill step, create/update task tracking for all planned steps, then keep it synchronized as each step starts/completes.
545
546## [LESSON-LEARNED-REMINDER] [BLOCKING] Task Planning & Continuous Improvement — MANDATORY. Do not skip.
547
548Break work into small tasks (task tracking) before starting. Add final task: "Analyze AI mistakes & lessons learned".
549
550**Extract lessons — ROOT CAUSE ONLY, not symptom fixes:**
551
5521. Name the FAILURE MODE (reasoning/assumption failure), not symptom — "assumed API existed without reading source" not "used wrong enum value".
5532. Generality test: does this failure mode apply to ≥3 contexts/codebases? If not, abstract one level up.
5543. Write as a universal rule — strip project-specific names/paths/classes. Useful on any codebase.
5554. Consolidate: multiple mistakes sharing one failure mode → ONE lesson.
5565. **Recurrence gate:** "Would this recur in future session WITHOUT this reminder?" — No → skip `$learn`.
5576. **Auto-fix gate:** "Could `$code-review`/`$code-simplifier`/`$security-review`/`$lint` catch this?" — Yes → improve review skill instead.
5587. BOTH gates pass → ask user to run `$learn`.
559 **[CRITICAL-THINKING-MINDSET]** Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act.
560 **Anti-hallucination principle:** Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.
561 **AI Attention principle (Primacy-Recency):** Put the 3 most critical rules at both top and bottom of long prompts/protocols so instruction adherence survives long context windows.
562 **Goal-driven execution:** Define success criteria first, loop until verified, and stop only when observable checks pass.
563 **Tests verify intent:** Tests must protect business rules/invariants and fail when the protected intent breaks, not only mirror current behavior.
564
565## Common AI Mistake Prevention (System Lessons)
566
567- **Re-read files after context compaction.** Edit requires prior Read in same context; compaction wipes read state. Re-read before editing.
568- **Grep for old terms after bulk replacements.** AI over-trusts find/replace completeness. Grep full repo after bulk edits for missed refs in docs/configs/catalogs.
569- **Check downstream references before deleting.** Deletions cascade doc/code staleness. Map referencing files before removal.
570- **After memory loss, check existing state before creating new.** Compaction wipes prior-work memory. Query current state to resume — never blindly duplicate.
571- **Verify AI-generated content against actual code.** AI hallucinates APIs, class names, method signatures. Grep to confirm existence before documenting/referencing.
572- **Trace full dependency chain after edits.** Changing a definition misses downstream consumers. Trace the full chain.
573- **When renaming, grep ALL consumer file types.** Some file types silently ignore missing refs (no compile error). Search code, templates, configs, generated files.
574- **Trace ALL code paths when verifying correctness.** Code existing ≠ code executing. Trace early exits, error branches, conditional skips — not just happy path.
575- **Update docs that embed canonical data when source changes.** Docs inlining derived data (workflows, schemas, configs) go stale silently. Update all embedding docs alongside source.
576- **Verify sub-agent results after context recovery.** Background agents may finish while parent compacted — grep-verify output, don't trust assumed completion.
577- **Cross-check full target list against sub-agent assignments.** Parallel sub-agents by category miss boundary items. Reconcile union of assignments against target list before proceeding.
578- **Sub-agents inherit knowledge only from their agent .md definition — use custom agent types, not built-in Explore.** Tool adoption = permission + knowledge + enforcement (numbered workflow step).
579- **Persist sub-agent findings incrementally, not as a final batch.** Long sub-agents hit cutoffs before final write — findings lost. Instruct append-per-section to report file.
580- **When debugging, ask "whose responsibility?" before fixing.** Trace caller (wrong data) vs callee (wrong handling). Fix at responsible layer — never patch symptom site.
581- **Test failure → record a provisional verdict before trace/edit, then investigate.** Use the full five-way taxonomy: SOURCE-WRONG (production violates intent), TEST-WRONG (assertion/setup is stale), TEST-NOT-OPTIMAL (valid but fragile or low-signal test), ENVIRONMENT-BLOCKED (external state prevents a verdict), or AMBIGUOUS (intent/evidence cannot choose safely). Then trace root cause and triangulate against the governing spec (`docs/specs/**` if one exists) AND source. NEVER weaken an assertion, add a skip, relax a timeout, or change source merely to force green.
582- **Grep ALL removed names after extraction/refactoring.** Primary file "done" ≠ secondary files clean. Grep entire scope for every removed symbol before declaring complete.
583- **Assume existing values are intentional — ask WHY before changing OR flagging one as a defect.** Pattern-matching as "wrong" skips context. Before changing or reporting any constant/limit/flag/cutoff: read comments, git blame, the CALLER's ordering (the guarantee that makes the value correct usually lives in code running immediately BEFORE the cited line), and 2+ sibling call sites of the same convention. A doc stating WHAT without WHY is missing rationale, not proof of a missing guard — and in a validation pass, an accurate `file:line` citation proves the transcription, never the defect.
584- **Verify ALL affected outputs, not just the first.** One build green ≠ all green. Multi-stack changes (backend/frontend/tests/docs) require verifying EVERY output.
585- **Evaluate fit before copying a nearby pattern.** Closest example ≠ matching preconditions — verify the new context shares the same constraints, base classes, scope, lifetime.
586- **Holistic-first debugging — resist nearest-attention trap.** Don't dive into first plausible cause. List EVERY precondition (config, env vars, paths, DB, endpoints, creds, versions, DI, data). Verify each against evidence (grep/query — not reasoning). Ask "what would falsify this?" — if nothing, it's not a hypothesis. Most expensive failure: going deeper in "obvious" layer while bug sits in layer never questioned.
587- **Surgical changes — apply the diff test (context-aware).** Two modes: (1) Bug fix → every line traces to the bug; no restyling; orphan cleanup only for imports YOUR changes made unused. (2) Review/enhancement → implement improvements AND announce as "Enhancement beyond main request: [what]". Never silently scope-creep. Diff test: "Would this line exist if I wasn't asked to do X?" — if no, delete or announce.
588- **Surface ambiguity before coding — don't pick silently.** Multiple valid interpretations → present each with effort: "[Request] could mean (1) [N h], (2) [N h]. Which matters?" List scope/format/volume/constraints assumptions first. If simpler path exists, say so. Never silently pick.
589- **[MANDATORY FIRST ACTION] ALWAYS activate a suitable skill or workflow BEFORE responding.** Match task against workflow catalog + skill list; invoke via skill invocation or `$start-workflow <workflowId>`. NEVER answer or write code before checking. Skip = protocol violation.
590- **Why-Review adversarial mindset — apply when reviewing any plan, decision, or design.** Default SKEPTIC not VALIDATOR: steel-man a rejected alternative, invert each stated reason ("what does it sacrifice?"), stress-test top 2-3 assumptions, run pre-mortem ("ships, fails in 3 months — what breaks?"), surface 1-2 alternatives author missed. Section presence ≠ quality; quality = causal reasoning + concrete mitigations + evidence, not "it's better" or "monitor closely".
591- **Front-load report-write in sub-agent prompts for large reviews.** Many-file sub-agents hit budget before final write — findings lost. Design prompts so: (1) report-write is first explicit deliverable, (2) append per-file/section (not batched), (3) scope bounded so reads don't exhaust budget. Truncated mid-sentence with no report file → spawn narrower scope, don't retry same prompt.
592- **After context compaction, re-verify all prior phase outcomes before continuing.** Summaries describe intent, not environment state (git index, filesystem, processes). On resume, FIRST audit: git status, re-read modified files, verify filesystem. Every "completed" claim is an untested hypothesis until evidence confirms.
593- **OOM/memory: check row count before row size.** Triage: (1) Unbounded query — no DB filter for trigger? Push filter to DB; eliminates OOM. (2) Large rows? Projection reduces proportionally. Row reduction > projection in ROI.
594- **Assert the outcome your system OWNS, never the intermediate state your INFRASTRUCTURE owns.** When testing anything asynchronous (queue/broker delivery, retries, background jobs, caches, replication), assert the final business/entity state. NEVER assert the delivery bookkeeping — consume/send status, attempt counts, last-error, row existence or counts in a broker, scheduler, or outbox/inbox table. That bookkeeping lives in shared infrastructure that ANY co-running process (a peer worker, a second replica, a leftover local container) can write, usually under a deterministic shared key, so the assertion silently tests the developer's environment instead of the system: green when run alone, flaky the instant anything else shares that broker + database. Gate question for every assertion: "would this hold no matter WHICH process did the work?" — if no, assert the converged data state instead. Corollary: process-local fault injection and in-process telemetry cannot gate work any process may perform — use them as stress amplifiers (arm → bounded window → disarm → assert convergence), never as preconditions.
595- **Keep domain concepts out of generic/shared/infrastructure layers.** Reusable layer (shared library, framework, infra module) must reference NO consumer-specific domain concept — tenant/customer/product IDs, business entities, feature rules. Leak compiles + runs → passes review silently while coupling the "reusable" layer to one consumer. Keep shared type domain-free; push domain fields/logic down into the consumer via subclass/composition. — why: a layer coupled to one consumer's domain is no longer reusable.
596
597<!-- CODEX:SYNC-PROMPT-PROTOCOLS:END -->
duc01226