メニュー
Create new skills or review and improve existing ones. Use when user says create skill, new skill, improve skill, review skill, audit skills, or skill quality.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Execution trace capture and token baselining. Use when user says trace, extract trace, count tokens, token baseline, or /harness-trace. Not for harness optimization (use harness-mechanic).
Release prep and tagging with pre-flight verification. Use when user says release, tag, ship it, push to production, create release, or bump version. Not for editing a version string outside a release, a lockfile or a changelog entry by hand, and not for reading what changed in a past release: those are ordinary edits, just do them.
Mine LEARNING.md retrospectives across all repos for recurring failure-modes, then propose harness changes (hooks, rules, skills, checklist items) with falsifiable change-contracts. Also runs the doc-gardening pass over the governance docs (stale rules, dead cross-references, skill-table drift). Use when user says learning loop, mine learnings, recurring failures, what should I fix in my process, cross-repo retrospective, doc gardening, stale docs, or run learning-loop. Runs on human schedule, never autonomously. Not for single-project retrospectives (use learning-docs) or trace/token analysis (use harness-mechanic).
Full contractor-mode loop: sub-protocols (LOCALIZE, BENCH-BASELINE, REPRODUCE, DRIFT), review routing by file pattern, blast radius, UAT, parallelism and effort caps, escalation, just-do-it, goal-backed runs. Load it to RUN that loop, once you have an approved plan and are entering step 1, or when a step needs its detail. It is a protocol, not a way to do the work: it never replaces the language, review or planning skill the task itself calls for, and a task small enough to just edit (SKIP_SET) needs none of it. The always-on spine lives in rules/orchestrator-protocol.md. Not for trace capture (harness-trace) or harness optimization (harness-mechanic).
Ruby on Rails with service-oriented architecture, Dry-validation, Sidekiq/Solid Queue, Hotwire. Use for Rails API, Rails services, Rails forms, RSpec, ActiveRecord, Rails migrations. Not for standalone Ruby gems (use ruby skill), and not for judging whether a migration is safe to run in production (that is the database-reviewer agent).
Get three independent second opinions (isolated Claude + isolated Gemini + isolated DeepSeek) on a problem Claude is analyzing. Use when user says second opinion, another perspective, challenge this approach, or asks gemini/deepseek about a decision, approach, or diagnosis. Also invoke on your own initiative, without being asked: on a complex task before planning, after 2 failed root-cause attempts, on a 🔴 decision (rewrite, core logic, security), or when the review loop is stuck below 80 after 2 fix rounds. Gathers context, writes a focused prompt, calls all reviewers in isolated Docker containers, and synthesizes all viewpoints. Not for reviewing code or changes, even phrased as 'ask gemini to review' (use gemini-review).
| name | skill-forge |
| description | Create new skills or review and improve existing ones. Use when user says create skill, new skill, improve skill, review skill, audit skills, or skill quality. |
Create new skills or audit/improve existing ones. Encodes Anthropic's official guide + our conventions.
/skill-forge create <name>/skill-forge review <name> or /skill-forge review all---
name: kebab-case-name
description: "[What it does]. Use when [trigger conditions]. [Key capabilities]. Not for [negative triggers if overlapping skills exist]."
compatibility: "Only if external deps needed"
---
# ABOUTME: [One-line what]
# ABOUTME: [One-line key capabilities]
# Skill Title
## Quality Notes (if multi-step workflow)
- Take your time with each step
- Quality over speed
- Do not skip validation
## [Core Instructions]
### Step 1: ...
[Specific, actionable instructions with examples]
### Step 2: ...
[Clear expected output]
## Common Issues
| Issue | Solution |
|-------|----------|
## References (if >150 lines needed)
For detailed patterns, consult `references/<topic>.md`
| Check | Rule |
|---|---|
| Folder name | kebab-case, no spaces/capitals/underscores |
| File name | Exactly SKILL.md (case-sensitive) |
| Frontmatter | --- delimiters, name + description required |
| ABOUTME | Two lines after frontmatter: what + capabilities |
| No README.md | All docs in SKILL.md or references/ |
| Check | Rule |
|---|---|
| Formula | [What] + [When/triggers] + [Capabilities] |
| Trigger phrases | Include words users actually say |
| Negative triggers | "Not for X" if overlapping skills exist |
| Length | Under 1024 characters |
| No XML tags | No < or > in description |
| Quotes | Wrapped in double quotes |
Test: Ask yourself "When would you use the [name] skill?" If the description doesn't answer clearly, it needs work.
| Check | Rule |
|---|---|
| Specific | Actionable commands, not "validate the data" |
| Examples | Concrete code/commands, not abstract descriptions |
| Error handling | Common issues table for workflow skills |
| Progressive disclosure | SKILL.md <150 lines core; detailed content in references/ |
| Quality notes | Multi-step workflows get anti-laziness section |
| Field | When to use |
|---|---|
compatibility | External deps (app, CLI, API key, MCP server) |
allowed-tools | Restrict to specific tools |
tools | Simpler tool restriction |
metadata | Distribution (author, version, mcp-server) |
Start at 100, subtract per finding:
| Severity | Deduction | Examples |
|---|---|---|
| Critical | -25 | No frontmatter, no description, wrong file name |
| Major | -10 | No trigger phrases, vague instructions, no ABOUTME |
| Minor | -3 | Missing compatibility, no error handling, no examples |
| Score | Verdict |
|---|---|
| 90+ | Ship it |
| 70-89 | Needs fixes |
| <70 | Rewrite |
_AST_GREP.md, _PATTERNS.md, source-control where relevantdescription: field, which is why it must state when to use the skill and when not to