Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Creates, refines, repairs, or decomposes Markdown plus Mermaid workflow diagrams (flowchart or stateDiagram) with explicit approval gates, staged writes, empirical Mermaid validation, and compact run reports. Use when generating process flows, Mermaid flowcharts or state diagrams, agent operating procedures, human-in-the-loop gate maps, or decomposing a skill package flow diagram.
Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me". Supports a depth selector (quick / standard / deep) — deep is the default.
Generate hierarchical AGENTS.md structure for codebases to optimize AI agent token usage. Use when creating AGENTS.md files, documenting codebase structure, setting up agent guidance, organizing project documentation for AI tools, implementing JIT indexing, or working with monorepos that need lightweight root guidance with detailed sub-folder documentation. Covers repository analysis, root AGENTS.md generation, sub-folder AGENTS.md creation, and token-efficient documentation patterns.
Initialize, migrate and incrementally update project AI memory files, with AGENTS.md as the single source of truth. Intelligently completes memory entries through interactive selection and diff comparison. Trigger conditions: invoke proactively when the user mentions initializing memory files, updating AGENTS.md, migrating CLAUDE.md/GEMINI.md, syncing AI memory, init-ai-md, or similar keywords. Use when initializing, migrating, or incrementally updating project AI memory documents. The skill maintains the canonical root AGENTS.md, redirects CLAUDE.md and GEMINI.md, and manages the project-level record-bug-fix-memory skill.
Initialize, update or overwrite the Claude Code status line configuration files (.claude/settings.json and .claude/statusline.sh). Use to quickly set up status line display in any project, showing directory, Git branch, model, version and context window usage.
Make knowledge-work projects self-describing to agents and resumable across sessions, and govern the project's rules and working discipline. Two use cases: (1) project organization — creating or initializing a project, tidying or health-checking a project structure, designing directories and navigation, writing frontmatter/index/log/decision records, settling temporary files from the exploration phase into durable form, restoring context for a project that has no rules file yet; (2) working methodology — crystallizing recurring issues into rules, trimming rules that are too numerous or conflicting, adding or removing a project rule, setting up check scripts or gates for a project and judging whether they work, distilling lessons learned, reviewing or improving the methodology itself. Triggers: new project, initialize doc structure, help me tidy this project, how should a project be organized, health check, hit that pitfall again, crystallize into rules, too many rules, trim rules, gate, check script, distill, lessons learned, decision record, review methodology, setup project, project health check.
Investigate, document, or change an unfamiliar legacy codebase using local repository maps, scoped source context, and verified evidence. Use for undocumented internal frameworks, cross-module behavior, or large repositories; a known-file edit or single-symbol search usually needs only direct search.
Periodic pass that keeps a project's verification skill and feature map honest: parallel source readers per feature, one live session driving every feature, at most one PR of proven corrections. Use for /maintain-verification-skill or "audit the verify skill".
Use Mandex (mx) to search and manage offline documentation packages for AI agents. Covers pulling docs from CDN, full-text search with BM25 + optional ONNX semantic reranking, building custom .mandex packages from markdown, project-level dependency sync, and multi-agent integration (Claude Code, Cursor, Codex). Use when: looking up library documentation offline, building searchable doc packages, setting up documentation for coding agents, or searching across project dependencies.
Give an AI agent an encrypted inbox and agent-to-agent messaging CLI. Use when agents need stable inbox addresses, durable encrypted threads, signed shared channels, async multi-agent handoffs, human approvals, device-key recovery, or JSON automation.
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
Search GitHub for MCP (Model Context Protocol) server tools and automatically install and configure them into Claude config files. Triggered when the user needs to install MCP tools. Workflow: search MCP projects on GitHub -> extract npx config -> add to ~/.claude.json -> handle API keys (if any).
A security testing methodology for the MCP (Model Context Protocol) protocol. Triggers when the target environment uses MCP Servers to integrate external tools, when the security of MCP tool descriptions needs to be assessed, or when testing the security boundaries of Agents invoking tools via MCP. Covers: tool description poisoning, rug pull attacks (dynamic tampering), instruction override (Shadow Tool), hidden instructions (ANSI/Unicode), cross-Server attacks, token theft, schema manipulation, and context overflow.
Execute prompts using Minimax models through Cline CLI in headless mode. Use when users explicitly request Minimax execution, when a workflow needs a Minimax seat, or when a cross-runner workflow wants a Minimax-backed perspective without leaving the current workspace.
Restore a prior business-state snapshot saved by /money-save, so you can continue from exactly where the last session left off. Loads the most recent snapshot for the current project (auto-detected from the working directory), or a specific one by index. Use when starting a new Claude Code conversation and the user wants to continue prior work. Triggered by: 'continue from last time', 'where did we leave off', 'pick up from before', 'resume', '接着上次', '续上', '之前的结论', '上次诊断到哪了'.
Panel of experts — consults multiple agents from the ecosystem in parallel for multi-perspective analysis of any topic. Activates personas, experts and technical agents simultaneously, each with its own unique perspective, and consolidates them into a final decision-oriented synthesis.
Expert guide for using n8n-mcp MCP tools effectively. Use when searching for nodes, validating configurations, accessing templates, managing workflows, or using any n8n-mcp tool. Provides tool selection guidance, parameter formats, and common patterns.
Documents the `npx skills` CLI for installing and updating agent skills, plus the project-level `skills-lock.json` lockfile. On start, detects a `skills-lock.json` without installed skills under `.claude/skills/` and proposes restoring them via `npx skills add … --copy -y` per the source recorded in the lockfile.
Run a project-manager dispatch loop over a GitHub backlog: triage and queue ready issues, claim each one, dispatch it to a parallel developer agent that returns a structured JSON report, review the results against GitHub, and drive accepted pull requests to landing — escalating to the maintainer only what genuinely needs a human decision. Ships the developer-agent operating template the loop injects into every dispatch (no custom agent types required). Use when asked to "work through the backlog", "batch-dispatch issues", "派发 issue 给开发 agent", or to stand up a multi-agent delivery loop in an ObjectStack app project. Do not use for authoring ObjectStack metadata (the domain skills cover that), for a single already-scoped change you can just make, or as a replacement for the project's own conventions file — that file always wins.
OpenClaw official documentation knowledge base. Use when users ask about OpenClaw topics such as installation, configuration, Gateway, channel connections like WhatsApp/Telegram/Discord, Sessions, Tools, Skills, Pi Agent, CLI, and troubleshooting. Check local documentation first by default, and consider supplementing with web search only when the documentation is insufficient.
Scan a repository for Opus-4.6-era patterns that break or degrade on Opus 4.7 — fixed-budget Extended Thinking parameters, retired model ID aliases, and prompts that assumed verbose default output or eager sub-agent delegation. Produces a categorized report with file:line references and migration actions. Triggers on: "opus 4.7 migration", "migrate to opus 4.7", "audit for opus 4.7", "opus 4.6 to 4.7", "scan for budget_tokens", "find retired model IDs", "adapt repo to opus 4.7". NOT for implementing migrations — this skill identifies candidates.
Use the public `orca` CLI to operate Orca-managed worktrees, folder contexts, terminals, repos, automations, worktree comments, and the browser embedded inside the Orca app. Use when the user says "$orca-cli", "use orca cli", "Orca worktree", "child worktree", "cardStatus", "spawn codex/claude in a worktree", "read/wait/send Orca terminal", "terminal send", "full handoff", "handover", "give this to another agent", "another worktree", "Orca browser", or "control the browser inside Orca". Prefer this over raw `git worktree`, ad hoc PTYs, Playwright, or Computer Use when the task touches Orca-managed state. Use Computer Use for browser windows, webviews, or desktop UI outside Orca's embedded browser.
A general-purpose agent team orchestrator that takes any topic or task, assembles a team on its own, and handles research, analysis, review, and delivery of results. Triggered by the keywords "팀으로 분석해줘", "에이전트 팀으로 조사해줘", "다각도로 검토해줘", "심층 분석 부탁해", "여러 관점으로 봐줘", "think-team", "think team". Use in any situation requiring complex judgment, research, or strategic decisions rather than a simple question.
Build with and use Pi, the minimal terminal coding harness. Use for installing Pi, configuring providers/models/settings/environment variables, creating Pi skills/extensions/packages/themes/prompt templates, embedding Pi through the SDK, integrating over RPC or JSON event streams, parsing sessions, running local models through the llama.cpp router, developing custom Pi providers and TUI components, or using ecosystem packages such as pi-subagents (delegation/orchestration), pi-mcp-adapter (MCP servers), pi-interview (interactive forms), and pi-web-access (web search, fetching, video understanding).
Creates implementation-ready plans and specs for complex changes without writing code. Use when the user asks for a plan/spec/roadmap, requests a detailed plan, or when a change spans multiple files/systems and needs discovery, sequencing, or architecture decisions.
Transforms workflow to use Manus-style persistent markdown files for planning, progress tracking, and knowledge storage. Use when starting complex tasks, multi-step projects, research tasks, or when the user mentions planning, organizing work, tracking progress, or wants structured output.
Adversarial cooperation loop for ONE ticket — player implements, verify reviews, every changed default turn is pushed to a draft PR, CI passes, and the PR is marked ready. Invoke only when the user explicitly asks for the player-coach loop. Do not use for an ordinary coding request. For a whole epic, use epic-runner instead.
Run external coding agents (Codex, Claude Code, OpenCode, native Grok Build, Gemini, Devin CLI) as independent reviewers, stateful repository researchers, or single-agent implementers. Use for multi-model opinions and code review, steerable Grok research, full-access delegation, long-running work, reattaching to delegated runs, or showing current subagents in the optional UI. Not for simple questions answerable directly from docs or the current codebase.
Apply to any non-trivial work, not just bulk work: edits, migrations, analyses, checks. Build the tool that does it or proves it (codemod, script, generator, or a skill your subagents follow) instead of working by hand. The tool is the artifact a reviewer can rerun
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.