Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Discover and compose geoscience skills for Python coding tasks. Use when choosing a geoscience library, finding a skill for a data format or method, or planning work across seismic, well logs, geological modelling, inversion, and other geoscience domains.
Import an existing skill from a GitHub repository URL into the current workspace. Use only when the user provides or references a GitHub URL/repo to fetch from (e.g., `/add-skill https://github.com/OpenHands/extensions/tree/main/skills/codereview` or "add the codereview skill from https://github.com/OpenHands/extensions/"). Handles fetching the skill files and placing them in .agents/skills/. This does not author new skills — to create a new skill from scratch (no source URL), use the skill-creator skill instead.
Monetization and retention Supervisor — routes and coordinates conversion, customer experience, retention, average order value, email, SMS, and messaging. Trigger words: 变现, monetization, 留存, retention, 提升收入, LTV提升, 变现策略, retention revenue, ltv growth, revenue optimization.
Add a custom Python tool to an AG2 `Agent` using the `@tool` decorator. Use when the user wants to give an Agent a new capability backed by Python code (API calls, DB queries, computations, file ops). Covers sync and async tools, parameter typing, Pydantic schema customisation, returning typed `Input` / `ToolResult` (text / data / images / binary), `final=True` early-exit, and dependency injection via `Context` / `Inject` / `Variable` / `Depends`.
Pause an AG2 `Agent` mid-run to collect human input via `context.input()`, or gate a tool call with `approval_required()` middleware. Use when the user wants the agent to ask for confirmation, request missing info (passwords, API keys, data), or have a human approve sensitive / irreversible / expensive tool calls (sending emails, deleting records, payments).
Persist agent state across runs, shape what the LLM sees per turn, and cap history to fit a context window. Covers `KnowledgeStore` (memory / sqlite / disk / redis), `KnowledgeConfig` (`store=`, `expose_tool=`, `write_event_log=`, `compact=`, `aggregate=`, `bootstrap=`) and its opt-out flags, aggregation strategies (`WorkingMemoryAggregate` with `prompt=` override, `ConversationSummaryAggregate`), assembly policies (`WorkingMemoryPolicy`, `EpisodicMemoryPolicy`, `ConversationPolicy`, `SlidingWindowPolicy`, `TokenBudgetPolicy`, `AlertPolicy`), compaction (`TailWindowCompact`, `SummarizeCompact`), and the lifecycle events (`AggregationStarted/Failed`, `CompactionStarted/Failed`, `EventLogFailed`). Use when the user wants the agent to remember between conversations, manage long histories, or control prompt assembly.
Single-agent recursion and parallel fan-out within one AG2 `Agent` — auto-injected `run_subtask` / `run_subtasks(parallel=True)` (opt in via `tasks=TaskConfig(...)`) for self-delegation, and `Agent.as_tool()` as a lightweight no-hub way to call one named agent from inside another. Use when a single coordinator wants to break work into its own sub-tasks, fan out concurrent sub-tasks, or invoke a specialist agent as a tool. Covers context flow, recursion safety, and `persistent_stream` for sub-task history. **For two or more agents actually collaborating with a registry, durable channels, governance, or turn-taking, use `ag2-network-quickstart` instead** — the network is the standard multi-agent pattern in AG2.
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to “open a website”, “fill out a form”, “click a button”, “take a screenshot”, “scrape data from a page”, “test this web app”, “login to a site”, “automate browser actions”, or any task requiring programmatic web interaction.
Use this skill when the user wants to operate Airtap or complete a request through a mobile app on an Airtap device. It lists receivers and models, creates Airtap tasks, monitors them locally by default, sends follow-up messages, cancels tasks, updates user location, and can mirror progress updates into OpenClaw when needed.
Organize durable agent research outputs as summaries, analysis, and evidence dossiers. Use when producing multi-layer research artifacts or coordinating research handoffs. Do not use this skill for unrelated requests; route to the nearest named specialist.
Configure AWS MCP servers for documentation search and API access. Use when setting up AWS MCP, configuring AWS documentation tools, troubleshooting MCP connectivity, or when user mentions aws-mcp, awsdocs, uvx setup, or MCP server configuration. Covers both Full AWS MCP Server (with uvx + credentials) and lightweight Documentation MCP (no auth required)
Task tracking system for agents via Backlog.md CLI. Use when creating deferred issues during implementation, filing audit findings, working assigned tasks, or managing project work. Optimized for agent workflows: structured issue filing, priority/labeling system, and task completion tracking.
Design Bolna graph agents: node-based voice flows with LLM, expression, unconditional and event edges, static low-latency nodes, silence repeat with escalation, per-node RAG, function tools, and real-time event injection via REST. Use when a single prompt is too fragile for a multi-step voice workflow — appointment booking, lead qualification, collections, IVR with LLM fallback, payment confirmation, or any flow with branches, compliance checkpoints, or external events.
Use when the user asks to "optimize CLAUDE.md", "create a new skill", "write a custom agent", "configure hooks", "manage context window", "set up MCP servers", "scaffold a skill package", "analyze token budget", "create subagents", "configure permissions", "set up worktrees", or "integrate Claude Code with editors". Covers Claude Code CLI mastery, skill authoring, context engineering, hooks automation, subagent creation, and development workflow optimization.
Single "start here" orchestrator for new cold email operators. Asks infrastructure-status questions first, then runs /icp-onboarding → /lead-magnet-brainstorm → /campaign-strategy in sequence, producing a consolidated campaign-plan.md. Ends with an interactive menu of next skills branched on whether the user already has domains/inboxes. Use as the very first invocation on any new repo checkout or when onboarding someone to cold email from zero.
Structured disagreement protocols that expose weaknesses, test assumptions, and generate alternatives. Use when stress-testing proposals, playing devil's advocate, challenging architectural decisions, or auditing assumptions before finalizing plans.
Creating high-quality agent skills following Claude's official best practices. Use when designing, implementing, or improving agent skills, including naming conventions, progressive disclosure patterns, description writing, and appropriate freedom levels. Helps ensure skills are concise, well-structured, and optimized for context efficiency.
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
Fetch and analyze Bolna call execution data, raw logs, transcripts, recordings, costs, statuses, hangup details, extracted data, agent history, and batch executions. Use for debugging failed calls, analytics export, CRM sync, webhook reconciliation, and call monitoring.
Snapshots/restores Ghostty Claude Code/Codex tabs across reboots. Use before Mac updates (保存终端会话/重启前备份终端), after restart to reopen original session IDs (恢复之前的窗口/restore my ghostty tabs), audit survivors, or periodic change-only snapshots (定期增量快照/没变化不记录). Also for unmanaged/legacy Codex account switches (额度耗尽换号/换号后恢复全部 Codex); Minions native_runtime uses Account Manager. Not for --resume, screenshots or tmux.
Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.
Control Herdr, a terminal multiplexer for coding agents. Use Herdr to inspect or control panes, tabs, workspaces, terminals, run commands, and or starting and monitoring background processes like dev servers. Use for subagent when the user or another skills explicitly asks for or requies subagents. Requires HERDR_ENV=1.
Use when the user asks to “create a hook,” “add a PreToolUse/PostToolUse/Stop hook,” “validate tool use,” “implement prompt-based hooks,” “use ${CLAUDE_PLUGIN_ROOT},” “set up event-driven automation,” “block dangerous commands,” or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks, focusing on the advanced prompt-based hooks API
Comprehensive SAP Joule CLI (formerly sapdas CLI) assistant for managing digital assistants from the command line — compiling capabilities, deploying assistants, running BDD tests, linting, and troubleshooting errors. Use this skill whenever the user mentions "joule cli", "sapdas", "joule compile", "joule deploy", "joule test", "joule login", "joule lint", digital assistant deployment, capability compilation, DAAR files, RTA artifacts, or any task involving the Joule command line interface — even if they just say something like "deploy my assistant" or "how do I log in to Joule from the terminal". Also trigger when the user asks about testing Joule capabilities with Cucumber, linking AI assistants, managing deployed assistants, or automating Joule workflows in CI/CD pipelines.
Build multi-agent AI systems with LangGraph — the low-level orchestration framework for stateful, graph-based agent workflows. Covers supervisor, swarm, and hierarchical multi-agent patterns; subgraph composition; state management (checkpointers/stores); persistence; evals; and production debugging. Reach for this when designing agent architectures that need cycles, conditional branching, parallel execution, or human-in-the-loop patterns. Do not use this skill for unrelated requests; route to the nearest named specialist.
Use this skill when the user says "升级 skills", "更新 li-skills", "li-skills 有新版本吗", "upgrade skills", or "skills 怎么更新". It should also trigger even if the user only asks "有新版本吗", and proactively run the upgrade process. Check the current version, run npx skills update or reinstall jiangjiax/li-skills, then prompt the user to restart the Claude Code session to load the new version. Use when the user wants to "upgrade skills", "update li-skills", or get the latest version of the li creator toolkit.
Use this skill when the user says "优化流程", "检查流程", "系统哪里有问题", "流程有漏洞", or "新增了 XX 功能，更新流程". Read documents such as CLAUDE.md and 工作流程.md, diagnose process gaps, identify automation opportunities, generate an optimization plan, and directly apply the changes, saving the optimization report to 04-方法论沉淀/. Should not trigger for: content creation tasks (use the corresponding li-* skill) or skill development (use li-factory). Use when the user wants to "optimize workflow", "check process", "improve system", or notices a gap in the content creation pipeline.
Persistent local memory for OpenClaw agents. Use when users say: - "install memos" - "install MemOS" - "setup memory" - "add memory plugin" - "openclaw memory" - "memos onboarding" - "memory not working" - "configure memory" - "enable memory" - "upgrade MemOS" - "update memory plugin"
Designs, builds, debugs, and documents n8n workflows and AI agent automations. Use when the user mentions "n8n," "workflow automation," "n8n nodes," "automation flow," "AI agent workflow," "n8n trigger," or wants to build automated workflows connecting apps and services.
Use when asked to fix, build, refactor, review, verify, or release software and the work is non-trivial — including delivering a change request through planning, gates, implementation, review, verified PR, and authorized release; continuing work started elsewhere with no packet via evidence-based position assessment; or operating tracked work items in Linear, Jira, or Notion via their tooling skills. neckbeard routes the change through framing, discovery, design, implementation, review, verification, and learning — choosing the smallest *safe* intervention, proving it at the real delivery boundary, and leaving an inspectable evidence ledger. For issue-to-PR work, conditionally loads a 9-phase journey with gates and lifecycle integration. Composes specialist catalog skills rather than replacing them. Not a persona or LOC-minimizer. Do not use this skill for unrelated requests; route to the nearest named specialist.
Invoke OpenCode CLI as a sub-agent to leverage alternative AI models (GPT-5.x, Codex, local models) for code implementation, review, debugging, research, and second opinions. Use when the user asks to "use opencode", "get a second opinion from GPT/Codex", "implement with opencode", "review with opencode", "let opencode handle this", "use a different model", or wants cross-model validation. Also triggers when the user explicitly names an OpenCode model (e.g., "use gpt-5.4", "use codex") or wants to delegate a task to a non-Claude AI system. Supports all providers configured in the user's OpenCode installation (OpenAI, GitHub Copilot, local models via Ollama/LM Studio, and OpenCode's own free-tier models).
Make an answer easier to start, scan, pause, and resume without making it shallower: a reply that is correct but too dense to enter, a wall of text with the useful part buried, a summary that dropped the caveat that mattered, or prose that reads as machine-generated. Covers useful-answer-first ordering, short paragraphs, literal headings, preserved nuance, and a small diagram where structure is the difficulty. Use when the user says "I do not understand", "explain this more simply", "make this clearer", "too dense", "break it into parts", "step by step", or "without losing the details". Not for shortening at the cost of accuracy, not for compressing an ordered procedure or a security warning, and not for a draft that sounds generated rather than dense, which is a question about voice and belongs elsewhere. An answer nobody can get into was not delivered, and one shortened until it is wrong was not either.
Maintain a project's canonical terms and settled decisions that future work should reuse, and record follow-ups a session discovers but does not resolve. Use when project-specific terms are being clarified, when a choice that may constrain future work is being considered or settles, including during planning, or when a session applies a temporary workaround whose root cause stays open or observes an out-of-scope defect with evidence. Do not use for simple definitions, routine implementation details, or carrying out an already-settled decision.
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.