audit-recon
Use when auditing cartographer recon-brief injections in the audit log. Reports (repo x day) inject uniqueness, missing-atlas repos, stale-at-inject rate, and excerpt-size distribution. Triggers on "audit recon", "recon audit", "recon injections".
catchup-ecc
Use when scanning new commits in the external affaan-m/everything-claude-code repo for ideas worth adopting. Triggers on "catchup ecc", "check ecc", "what is new in ecc", "check affaan".
brush
Use when running a project hygiene pass on a repo - git cleanup, leftover triage, evolution check, backlog review, AGENTS.md refresh. Triggers on "brush", "brush the project", "project hygiene", "git hygiene", "apply git hygiene", "tidy repo", "clean up this repo".
design-agent-loop
Use when designing a goal-oriented agent loop or reviewing one for runaway risk - build gate, machine-decidable goal, anti-Goodhart boundaries, five failure modes. Triggers on "design a loop", "review this loop", "will this loop run away".
explain-interactively
Use when turning a codebase into an interactive HTML course for non-technical learners. Triggers on "turn this into a course", "explain interactively", "codebase walkthrough", "teach this code".
review-design-doc
Use when reviewing a system design document interactively (provide doc path). Walks findings one-at-a-time with three resolution options each; edits applied directly to the doc. Triggers on "review design doc", "critique design", "design review".
spike
Use when an idea is too fuzzy to spec confidently and a throwaway build would answer faster than more spec work - runs before elicit-requirements or create-prd. Triggers on "spike", "spike this", "prototype this idea", "throwaway prototype".
run-autopilot
Use when running a PRD end-to-end autonomously - catchup, plan, work, and a review-rework loop running consensus, blind, and doubt lenses every cycle. Triggers on "autopilot", "run autopilot", "autopilot status", "drain backlog".
review-work-completion
Use after all tasks are completed to validate implementation against PRD requirements via multi-agent consensus review; for a spec-only blind lens use review-blindly. Triggers on "review work", "check completed work", "are we done".
design-solution
Use when turning a PRD into a reviewed design doc (the HOW) before planning tasks - architecture fit, exact interfaces, reuse, alternatives, then an autonomous adversarial review. Triggers on "design solution", "design this PRD".
review-discovery-doc
Use when reviewing a feature discovery document before create-prd, interactively for completeness, coherence, integrity, right-sizing, and evolvability. Triggers on "review discovery doc", "critique discovery", "review discovery".
review-prd-backlog
Use to review the PRD backlog (dev/local/prds/backlog) before /run-autopilot - create-prd compliance, collisions, sizing/regrouping, gaps, goal alignment. Triggers on "review backlog", "review prds", "audit backlog", "backlog ready".
digest-github-repo
Use when generating a zettelkasten digest of GitHub repo activity (issues, PRs, commits) for triage. Two modes (curated-list for awesome-lists, activity-digest for general repos). Triggers on "digest REPO", "what is new in REPO", pasted GitHub URL.
elicit-requirements
Use when fleshing out a rough idea before /create-prd, when requirements are unclear or complex (adaptive depth, minimal to comprehensive). Triggers on "elicit requirements", "flesh out this idea", "what should we build", "scope this out".
extract-plugin
Use when extracting a cluster of personal skills/hooks/commands into a new Claude Code plugin repo and publishing to the buvis marketplace. Triggers on "extract plugin", "ship as plugin", "make plugin from", "spin out plugin".
manage-agents-md
Use when creating, auditing, or improving an AGENTS.md file, or promoting project memories into it (progressive disclosure, context budgeting). Triggers on "agents.md", "AGENTS.md", "audit AGENTS.md", "promote memories", "memories to AGENTS.md".
sort-agent-setup
Use when a project needs a trimmed Claude install instead of the full global setup, sorting skills/rules/hooks into DAILY vs LIBRARY with repo evidence. Triggers on "agent sort", "trim skills for repo", "daily vs library", "project install plan".
purge-devlocal
Use when purging stale dev/local artifacts across repos (review debris, done-PRD satellites, old autopilot state) via trash-first GC. Triggers on "purge dev local", "clean dev local", "gc dev local", "empty dev local trash".
research
Use when researching a topic via multi-source synthesis (literature review, competitive analysis, tech comparison); for a fact-checked cited web report use deep-research. Triggers on "research this", "investigate", "survey the landscape".
review-blindly
PRD-only blind lens. Use when a major implementation is complete and per-task reviews passed, before merge/handoff; for consensus validation use review-work-completion. Mandatory after multi-file features. Not for single-file fixes.
brief-portfolio
Use when the user wants a portfolio-wide status brief of all gita-registered repos as a single-file HTML dashboard with actionable follow-ups. Triggers on "brief portfolio", "state of my repos", "repo dashboard", "cross-repo todos".
work
Use when executing already-planned tasks one at a time, dispatching the implementor and committing after each. Triggers on "work on tasks", "implement tasks", "start working", "execute the plan", "do the work".
use-sonnet
Use when running Anthropic Claude Sonnet via the native claude CLI (headless) for code analysis, refactoring, or editing. Triggers on "run sonnet", "sonnet analyze", "ask sonnet".
use-qwen
Use when running a local Qwen model (llama.cpp + pi agent) for code tasks, or onboarding, qualifying, or promoting a new local model id. Triggers on "run qwen", "qwen analyze", "ask qwen", "onboard model", "qualify model", "promote model".
survey
Use to build or refresh the Cartographer per-repo atlas via /survey - a small codebase map of layers, naming conventions, error style, extension points. Triggers on "/survey", "survey repo", "refresh atlas", "rebuild codebase map".
claude-in-new-window
tmux の新しい window を作成し、指定ディレクトリで claude を起動して、指定した指示文を送信する。以下の場合に発動: (1) ユーザーが /claude-in-new-window <dir> <task> と入力した時、(2) 他 Skill から新 window で claude に作業を委譲する目的でチェイン呼び出しされた時、(3) PostToolUse hook (on-worktree-add.sh) が git worktree add を検知して間接的に呼び出す時。tmux 外では動作しない。現在の pane はそのまま元のディレクトリに留まる。
save-knowledge
現在のセッションでの Q&A を、別リポジトリで作業中でも muryoimpl/knowledge-docs リポジトリにエントリとして保存し INDEX を再生成する。「この内容を保存」「ナレッジに残して」「/save-knowledge」などで起動。
refactorer
コード品質の問題を積極的に検出し改善する。以下を検出した際に自動発動: (1) 長いメソッド/クラス、(2) 重複コードパターン、(3) 深いネスト、(4) 不適切な命名、(5) マジックナンバー/文字列、(6) N+1 クエリパターン、(7) 未使用コード、(8) Feature Envy や God クラス。言語非依存のリファクタリングガイダンスを提供。
shortcuts
Claude Code のキーボードショートカット一覧を表示する。ユーザーが /shortcuts と入力した時に発動。
committer
Git コミットの作成を支援する。以下の場合に発動: (1) ユーザーが /commit と入力した時、(2) コード変更が完了しコミットが適切と判断した時、(3) ユーザーがコミットを依頼した時。変更内容を分析し、適切なコミットメッセージを生成してコミットを作成する。
frontend-design
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifacts, posters, or applications (examples include websites, landing pages, dashboards, React components, HTML/CSS layouts, or when styling/beautifying any web UI). Generates creative, polished code and UI design that avoids generic AI aesthetics.
frontend-dev
Comprehensive React, TypeScript, and Tailwind frontend development with Next.js and TanStack ecosystem (TanStack Router, TanStack Query, TanStack Start). Use when building, configuring, or developing modern web applications with these specific technologies, including project setup, routing, data fetching, component patterns, and TypeScript configurations.
skill-creator
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
template-skill
Replace with description of the skill and when Claude should use it.
review-pr
Post a senior-staff-engineer-level code review with inline comments directly on a GitHub PR. Requires a PR URL or number as argument. Use when the user provides a GitHub PR link, says '/review-pr <number>', or asks to review a remote/existing PR on GitHub. Do NOT use for reviewing local uncommitted changes (that is ce-review).
codebase-memory
Use the codebase knowledge graph for structural code queries. Triggers on: explore the codebase, understand the architecture, what functions exist, show me the structure, who calls this function, what does X call, trace the call chain, find callers of, show dependencies, impact analysis, dead code, unused functions, high fan-out, refactor candidates, code quality audit, graph query syntax, Cypher query examples, edge types, how to use search_graph.
wechat-article-formatting
Convert Markdown content to WeChat Official Account HTML format with inline styles. Generates complete HTML preview files that can be copied directly to WeChat editor. Use when user needs to format articles for WeChat Official Account or mentions converting Markdown to WeChat HTML.
adhd-advisor
>-
api-engineering
>-
software-engineering
>-
secops-engineering
>-
routing-engineering
>-
quality-engineering
>-
php-assistant
>-
performance-engineering
>-
password-generator
>-
git-workflow
>-
frontend-engineering
>-
documentation-writer
>-
devops-engineering
>-
Page 1354 of 1713 · 85634 results
