test-driven-development
Use when implementing any feature or bugfix, adding tests, fixing flaky tests, refactoring, or changing behavior. Default approach for new features, bug fixes. Exceptions only for throwaway prototypes or generated code. Covers TDD workflow (red-green-refactor), condition-based waiting for async tests, and testing anti-patterns to avoid.
using-tmux
Control interactive CLI tools that require a real terminal via tmux send-keys. Use for git rebase -i, git add -p, vim, nano, Python/Node REPLs, or any command that hangs waiting for terminal input.
verification-before-completion
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
receiving-code-review
Process external code review feedback systematically. Use when responding to PR comments, review feedback, or technical suggestions from reviewers.
git-workflow
Use when starting feature work that needs isolation, or when completing work and deciding how to integrate - covers full worktree lifecycle from setup through merge/PR/discard
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior (including race conditions, deadlocks, concurrency issues) - four-phase framework (root cause investigation, pattern analysis, hypothesis testing, implementation) with specialized techniques for deep call stack tracing and concurrency debugging
agent-creator
Create custom Task subagents in ~/.claude/agents/. Use when defining new agents for Task tool delegation.
batch-operations
Use when performing 3+ similar operations like file edits, searches, or git operations. Triggers on repetitive tasks, "do the same for", loop-like patterns. Enables parallel tool calls and batch-editor delegation.
browser-automation
Browser automation via Puppeteer MCP for JS-rendered content
command-creator
Create slash commands for user-initiated workflows. Use when building repeatable /name commands for Claude Code.
debate-hall-admin
Admin operations for debate safety overrides and redaction.
debate-hall-auto
Automated Wind/Wall/Door debate orchestration with tier-based model selection.
debate-hall-github
GitHub integration for debate synchronization and ADR ratification.
debate-hall-manual
Manual turn-by-turn debate control for custom orchestration scenarios.
debate-hall
Wind/Wall/Door debate orchestration router. Routes to focused skills based on task.
instruction-improver
Search memex for user feedback patterns (frustration, corrections, praise, successful outcomes) to identify recurring mistakes and wins, then generate CLAUDE.md or AGENTS.md improvements.
memex-search
Search, filter, and retrieve Claude/Codex history indexed by the memex CLI. Use when the user wants to index history, run lexical/semantic/hybrid search, fetch full transcripts, or produce LLM-friendly JSON output for RAG.
recursive-knowledge
Process large document corpora (1000+ docs, millions of tokens) through knowledge graph construction and stateful multi-hop reasoning. Use when (1) User provides a large corpus exceeding context limits, (2) Questions require connections across multiple documents, (3) Multi-hop reasoning needed for complex queries, (4) User wants persistent queryable knowledge from documents. Replaces brute-force document stuffing with intelligent graph traversal.
research
Use when needing to understand requirements before implementation. Gathers context from Jira, Confluence, codebase, and docs. Produces research document with confidence scoring.
rpi
Use when implementing features from Jira tickets, PRDs, or user requirements. Orchestrates Research-Plan-Implement workflow with quality gates for hallucination, overengineering, and underengineering detection.
plan
Creates detailed implementation plan from validated research. Produces task breakdown with dependencies.
implement
Executes implementation plan with quality checks and progress tracking. Follows AGENTS.md patterns strictly.
audit-performance
Performance-focused audit that can run in background during implementation. Checks for inefficiencies, memory leaks, widget rebuilds. Injects P0 findings to main agent.
audit-security
Security-focused audit that can run in background during implementation. Checks for vulnerabilities, auth issues, data exposure. Injects P0 findings to main agent.
audit
Validates research/plan/code against overengineering, underengineering, and hallucination
code-review
Code reviewer focusing on correctness, regressions, security, and test coverage - P0/P1/P2 severity
agents-md-manager
Build and maintain effective AGENTS.md files that improve AI coding agent efficiency. Use when creating a new AGENTS.md file, auditing/improving an existing one, setting up progressive disclosure with agent documentation, or optimizing context window usage for AI-assisted coding projects.
catchup
Review branch changes since fork point from main/master. Use when resuming work on a branch or onboarding to existing changes. Triggers on "catch up", "catchup", "what changed on this branch", "summarize branch changes", "review branch".
inventory-notifications
Use when user wants to inventory, catalog, or summarize repositories from GitHub notifications for a specific repo. Triggers on "inventory notifications", "catalog repos from notifications", "summarize notification repos", "list repos from notifications".
plan-tasks
Creates granular implementation tasks from PRD documents. Use when user wants to plan work, create tasks from a PRD, implement a feature spec, or break down requirements into actionable steps. Triggers on phrases like "plan tasks", "create tasks from PRD", "implement PRD", "break down the spec", or when user wants to start working on a PRD document.
resolve-git-conflicts
Resolve git merge conflicts, rebase conflicts, cherry-pick conflicts, and stash pop conflicts. Use when git status shows "unmerged paths", "rebase in progress", "cherry-pick in progress", or after failed stash pop. Handles overlapping edits, edit/delete conflicts, rename conflicts, binary conflicts, and helps user choose between ours/theirs or manual merge.
restore-tasks
Restore tasks from a previous Claude Code session into the current session. Use when user wants to recover, restore, or load tasks from a past session, or when they mention tasks were lost after /clear or session restart. Triggers on "restore tasks", "load tasks", "recover tasks", "get my tasks back", "tasks from previous session".
review-plan
Review and critique the current plan created in plan mode. Use after creating a plan to identify gaps, risks, and get clarifying questions. Triggers on "review plan", "critique plan", "check my plan", or when user wants feedback on their implementation plan.
review-work-completion
Multi-agent review of completed work against PRD requirements. Use after all tasks are completed to validate implementation quality. Triggers on "review work", "check completed work", "validate implementation", "are we done".
save-plan
Save current plan into PRD file. Use after completing plan mode to persist the plan for later implementation. Triggers on "save plan", "save this plan", "create PRD from plan", "persist plan".
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.
sync-plan-issue
Create or update GitHub issue with current plan summary. Lists open issues for selection or creates new. Triggers on "sync plan to github", "update issue with plan", "share plan", "create issue from plan". Use after planning is complete and ready to share publicly. Do NOT use when plan is still in draft/exploration phase, working on private/sensitive features (issues are public), or issue has manually curated description to preserve.
use-codex
Use when the user asks to run Codex CLI (codex exec, codex resume) or references OpenAI Codex for code analysis, refactoring, or automated editing. Triggers on "run codex", "codex analyze", "use codex for", "codex refactor", "ask codex", or when user wants OpenAI's Codex model for code tasks.
use-gemini
Use when the user asks to run Gemini via copilot CLI for code analysis, refactoring, or automated editing. Triggers on "run gemini", "gemini analyze", "use gemini for", "ask gemini", "copilot with gemini", or when user wants Google's Gemini model for code tasks.
work
Implement pending tasks using Codex or Gemini, committing after each task. Use when ready to execute planned work. Triggers on "work on tasks", "implement tasks", "start working", "execute the plan", "do the work".
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。
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。
ask-questions-if-underspecified
Clarify requirements before implementing. Do not use automatically, only when invoked explicitly.
d2-diagram
Comprehensive tool for creating D2 diagrams based on descriptions and requirements. This skill should be used when creating visual diagrams, system architectures, flowcharts, network topologies, data flows, or any visual representation that can be expressed as a diagram. Triggers include requests to "create a diagram," "visualize," "draw an architecture," "show relationships," or when converting textual descriptions into visual diagrams.
pr-review
Perform thorough, constructive pull request reviews using parallel specialized agents. Use when user wants to review a PR, provide code review feedback, or assess code changes. Features confidence-scored issues, validation filtering, and batched GitHub comments.
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 Claude's capabilities with specialized knowledge, workflows, or tool integrations.
effect-best-practices
Enforces Effect-TS patterns for services, errors, layers, and atoms. Use when writing code with Effect.Service, Schema.TaggedError, Layer composition, or effect-atom React components.
effect-testing
Comprehensive testing patterns for Effect-TS services, errors, layers, and effects. Use this skill when writing tests for Effect-based code.
effect-ts
This skill should be used when the user asks about Effect-TS patterns, services, layers, error handling, service composition, or writing/refactoring code that imports from 'effect'. Also covers Effect + Next.js integration with @prb/effect-next.
Page 1223 of 1444 · 72197 results