Skills
96,770 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Generates custom design system rules for the user's codebase. Use when the user says “create design system rules,” “generate rules for my project,” “set up design rules,” “customize design system guidelines,” or wants to establish project-specific conventions for Figma-to-code workflows. Requires Figma MCP server connection.
Control tmux panes and communicate between AI agents. Use this skill whenever the user mentions tmux panes, cross-pane communication, sending messages to other agents, reading other panes, managing tmux sessions, or interacting with processes running in tmux. Includes tmux-bridge CLI for agent-to-agent messaging and raw tmux commands for direct session control.
Routes broad or ambiguous Sent requests to the correct MCP-backed operation or specialist skill. Use when the user asks what Sent can do, says “help me with Sent” or “set up messaging,” needs several Sent workflows, or has not made the channel, task, or desired operation clear enough to select a more specific skill.
Owns the session handoff lifecycle: the end-of-session ritual that captures completed work, pending tasks, and learnings into .claude/handoff.md, and the session-start protocol that loads it back. Triggers on: /wrap-up, "wrap up", "done for today", "that's all", "end session", "signing off", "handoff" — and at session start: "start session", "session start", "load handoff", "pick up where we left off", "what were we working on."
Interactive project initialization. Detects project type, asks architecture questions, and generates a customized CLAUDE.md — no manual template copying. Use when: "init project", "setup project", "initialize", "new project setup", "generate CLAUDE.md", "configure for dotnet-claude-kit", or starting Claude Code on a fresh or existing .NET codebase.
Mid-session save point: create a descriptive git commit and a brief handoff note, then keep working. Use before a risky change or refactor, when switching tasks, or to bank progress without ending the session. Triggers on: /checkpoint, "checkpoint", "save progress", "commit and handoff", "save state", "pause here", "before a risky change". For the full end-of-session ritual with learning extraction, use /wrap-up instead.
Turn a vague feature or product idea into an agreed, persisted specification through relentless structured questioning. Never assumes — every gap, ambiguity, or "probably" becomes a question to the developer, and the spec cannot be approved while open questions remain. Produces docs/specs/<NNN>-<slug>.md with acceptance criteria that /plan, /scaffold, and /tdd consume. Use when: "spec", "write a spec", "spec this out", "requirements", "PRD", "acceptance criteria", "define the feature", "user stories", "what should we build", or before planning any feature too big to describe in one sentence.
Enter plan mode for .NET projects with architecture awareness. Analyzes tasks through the lens of supported architectures (VSA, Clean Architecture, DDD, Modular Monolith) and produces structured implementation plans before any code is written. Use when: "plan", "let's plan", "think through", "design this", "how should I implement", or any non-trivial task requiring 3+ steps.
Use when working on complex multi-step tasks, when a session is getting long (40+ tool calls), when the agent starts ignoring rules it followed earlier, when conventions drift, when output quality seems to degrade, or after any context compaction event. Prevents long-session corruption AND context compaction amnesia through behavioral self-enforcement.
Use to configure, set up, or repair the Sales Management agent and Agentforce Pipeline Management in a Salesforce org. Automates metadata creation for flows, prompt templates, permission sets, and data source configuration. TRIGGER when: user wants to enable Pipeline Management, configure Sales pipeline features, set up the Sales Management agent for opportunity field updates (including autonomous updates), connect enabled data sources like Einstein Conversation Insights or Einstein Activity Capture, customize opportunity stage descriptions, configure post-meeting suggestions, verify or audit configuration status, fix partially configured orgs, or troubleshoot Pipeline Management metadata issues. DO NOT TRIGGER when: user wants to build a custom agent (use agentforce-generate), configure general Agentforce tracing (use platform-tracing-agentforce-configure), work with non-Sales agents, or enable Einstein Conversation Insights or Einstein Activity Capture from scratch (provisioning is out of scope).
Guide users through creating a new plugin from scratch in a cowork session. Use when users want to create a plugin, build a plugin, make a new plugin, develop a plugin, scaffold a plugin, start a plugin from scratch, or design a plugin. This skill requires Cowork mode with access to the outputs directory for delivering the final .plugin file.
Browser automation CLI for AI agents. Use when the user needs to inspect, test, or automate browser behavior: navigating pages, filling forms, clicking buttons, taking screenshots, extracting page data, reading selected OpenDesign browser-tab context, testing web apps, dogfooding OpenDesign previews, QA, bug hunts, or reviewing app quality. Prefer local OpenDesign preview URLs unless the user explicitly asks for external browsing.
A better way to find skills: identify the highest-quality Skill for your task using real-world feedback such as measured results and user reviews, rather than relying only on popularity metrics like download counts. Use this skill in either of these cases: ① The user wants to find, search for, recommend, or install a Skill, or learn from others’ experience—for example, “find me an xxx skill,” “how do others do stock analysis,” or “see if there’s an existing skill,” or they directly ask to find a skill / skill search. ② The Agent independently determines that an external Skill could help—for an unfamiliar task, or when prior attempts have failed and suitable tools are lacking. It can proactively look up practical experience and find matching Skills without waiting for the user to ask.
Customize a Claude Code plugin for a specific organization's tools and workflows. Use when: customize plugin, set up plugin, configure plugin, tailor plugin, adjust plugin settings, customize plugin connectors, customize plugin skill, tweak plugin, modify plugin configuration.
DeepEval evaluation workflow for AI agents and LLM applications. TRIGGER when the user wants to evaluate or improve an AI agent, tool-using workflow, multi-turn chatbot, RAG pipeline, or LLM app; add evals; generate datasets or goldens; use deepeval generate; use deepeval test run; send results to Confident AI; monitor production; run online evals; inspect traces; or iterate on prompts, tools, retrieval, or agent behavior from eval failures. AI agents are the primary use case. Covers Python SDK, pytest eval suites, CLI generation, traced evals, Confident AI reporting, and agent-driven improvement loops. DO NOT TRIGGER for unrelated generic pytest, non-AI test setup, or non-DeepEval observability work unless the user asks to compare or migrate to DeepEval; for instrumenting an app with DeepEval tracing, @observe, or framework integrations (use the `deepeval-tracing` skill); or for raw OpenTelemetry / OTLP export without the deepeval package (use the `deepeval-otel` skill).
Help the user define a concrete, measurable goal before starting work, especially when they ask to use the goal tool, create a goal, set an objective, clarify success criteria, or turn a fuzzy intention into a quantitative outcome. Use this skill for goal creation and goal refinement only; it does not manage durable snapshots, decision logs, or long-running execution artifacts.
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.
Delegate a coding task to the Warp Agent CLI (`oz`) as a background implementer, then review its diff and land it yourself. Use this whenever the user wants to hand implementation work to Warp — phrasings like "have Warp implement X", "delegate this to the Warp CLI", "run it through Warp", "use oz to implement/fix/refactor" — or wants to run a queue of coding tasks through Warp while staying the reviewer. DO NOT USE for tasks small enough to do inline, when the user wants the code written directly without delegating, or for the interactive `warp` TUI (this skill drives the headless `oz agent run`, not the terminal app).
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy
Use this when someone says "I don't know where to start" or "help me get going", asks "where am I?" in the six-stage journey, requests on-demand org feature detection for Data 360, OmniStudio, or DevOps Center, or asks what could help me do X / which plugin would help with X for a task no installed skill already covers. DO NOT TRIGGER for specific Salesforce tasks already owned by a leaf skill or for generic non-Salesforce help.
Search and browse the x402 bazaar marketplace for paid API services. Use when you or the user want to find available services, see what's available, discover APIs, or need an external service to accomplish a task. Also use as a fallback when no other skill clearly matches — search the bazaar to see if a paid service exists. Covers "what can I do?", "find me an API for...", "what services are available?", "search for...", "browse the bazaar".
Delegate a coding task to the Command Code CLI (`cmd`) as a background implementer, then review its diff and land it yourself. Use this whenever the user wants to hand implementation work to Command Code — phrasings like "have Command Code do X", "delegate this to commandcode", "run it through cmd", or "use Command Code to implement/fix/refactor" — or to run a queue of coding tasks through Command Code while staying the reviewer. DO NOT USE for tasks small enough to do inline, or when the user wants the code written directly without delegating.
Use when building custom agent backends, implementing the AG-UI protocol, debugging streaming issues, or understanding how agents communicate with frontends. Covers event types, SSE transport, AbstractAgent/HttpAgent patterns, state synchronization, tool calls, and human-in-the-loop flows.
@copilotkit/react-core — mount the CopilotKit provider (from @copilotkit/react-core/v2) in a Next.js App Router / React Router v7 / TanStack Start / SPA app, drop in CopilotChat/CopilotPopup/CopilotSidebar (v2 chat components ship from react-core/v2 — NOT react-ui, which is CSS-only in v2), access and subscribe to agents with useAgent / useAgentContext / useCapabilities, switch between multiple agents, manage durable Intelligence threads with useThreads, register browser-side tools via useFrontendTool, render tool calls with useRenderTool / useComponent / useDefaultRenderTool, gate execution with useHumanInTheLoop, wire file attachments with useAttachments, configure suggestion pills, and register activity- and custom-message renderers. Load the reference under references/ that matches your task.
Self-improving browser automation via the auto-research loop. Iteratively runs a browsing task, reads the trace, and improves the navigation skill (strategy.md) until it reliably passes. Supports parallel runs across multiple tasks using sub-agents. Use when you want to build or improve browser automation skills for specific website tasks.
Answer questions about the AI SDK and help build AI-powered features. Use when developers: (1) Ask about AI SDK functions like generateText, streamText, ToolLoopAgent, embed, or tools, (2) Want to build AI agents, chatbots, RAG systems, or text generation features, (3) Have questions about AI providers (OpenAI, Anthropic, Google, etc.), streaming, tool calling, structured output, or embeddings, (4) Use React hooks like useChat or useCompletion. Triggers on: "AI SDK", "Vercel AI SDK", "generateText", "streamText", "add AI to my app", "build an agent", "tool calling", "structured output", "useChat".
@copilotkit/runtime — mount a fetch-native CopilotRuntime on any JS server, wire middleware, pick an AgentRunner, instantiate BuiltInAgent (Factory Mode with TanStack AI is the preferred default) or plug in any of 12 external agent frameworks (Mastra, LangGraph, CrewAI Crews/Flows, PydanticAI, ADK, LlamaIndex, Agno, AWS Strands, MS Agent Framework, AG2, A2A), enable Intelligence mode for durable threads + websocket, register server-side tools via defineTool, and wire voice transcription. Uses the fetch-based createCopilotRuntimeHandler primitive — the Express/Hono adapters are discouraged. Load the reference under references/ that matches your task.
A skill that captures experience automatically and evolves safely. Triggers: (1) when a command or operation fails → record in ERRORS.md (2) when the user corrects you ('that's wrong'/'it should be') → record in LEARNINGS.md (3) when the user needs a capability that does not exist → record in FEATURE_REQUESTS.md (4) when an external API or tool errors → record in ERRORS.md (5) when you find your own knowledge is outdated or wrong → record in LEARNINGS.md (6) when you find a better approach → record in LEARNINGS.md (7) when each task completes → review the process and record any new experience in LEARNINGS.md. Deduplication rule: skip writing if there is no new experience or an existing entry already covers it. Every write also appends a JSONL log to .learnings/CHANGELOG.md. When an experience recurs 3 or more times, promote it to AGENTS.md/TOOLS.md/SOUL.md. See the body for details.
A cross-cutting cognitive mode for sitting with design problems before rushing to solve them. Part of the Intent design strategy system. Activates expansive brainstorming: hyperassociativity, beginner's mind, cross-domain pattern recognition, and suppression of premature idea-dismissal. Works alongside every Intent skill — strategize uses it to reframe briefs, blueprint to question structural assumptions, journey to rethink interaction models, and specify to stress-test specs. Trigger when the user invokes "expansive mode", "philosopher mode", "sit with this", "brainstorm", "explore this problem", or says things like "go weird with it", "don't filter yourself", "what connections are you not making", "think about this differently", or "I'm stuck". This is a reasoning protocol, not a character — Claude's voice stays grounded but the cognitive process changes significantly
Workflow orchestration for complex coding tasks. Use for ANY non-trivial task (3+ steps or architectural decisions) to enforce planning, subagent strategy, self-improvement, verification, elegance, and autonomous bug fixing. Triggers: multi-step implementation, bug fixes, refactoring, architectural changes, or any task requiring structured execution.
Use when adding CopilotKit to an existing project or bootstrapping a new CopilotKit project from scratch. Covers framework detection, package installation, runtime wiring (managed Intelligence or self-hosted SSE), provider setup, and first working chat integration.
Use when creating or maintaining OpenCLI site sitemaps: agent-facing navigation, page-state, action, workflow, API-reference, pitfall, and fallback knowledge for a website. Use after browser exploration discovers durable site context, when a sitemap is stale, or when promoting local site knowledge into the repo
CRITICAL: Use for ALL Rust questions including errors, design, and coding. HIGHEST PRIORITY for: 比较, 对比, compare, vs, versus, 区别, difference, 最佳实践, best practice, tokio vs, async-std vs, 比较 tokio, 比较 async, Triggers on: Rust, cargo, rustc, crate, Cargo.toml, 意图分析, 问题分析, 语义分析, analyze intent, question analysis, compile error, borrow error, lifetime error, ownership error, type error, trait error, value moved, cannot borrow, does not live long enough, mismatched types, not satisfied, E0382, E0597, E0277, E0308, E0499, E0502, E0596, async, await, Send, Sync, tokio, concurrency, error handling, 编译错误, compile error, 所有权, ownership, 借用, borrow, 生命周期, lifetime, 类型错误, type error, 异步, async, 并发, concurrency, 错误处理, error handling, 问题, problem, question, 怎么用, how to use, 如何, how to, 为什么, why, 什么是, what is, 帮我写, help me write, 实现, implement, 解释, explain
Discover, compare, and optionally install published Agent Skills from QianWen SkillHubs. TRIGGER when: user asks to find or recommend a skill, compare candidates, install a discovered skill, or explicitly invokes this skill by name. DO NOT TRIGGER when: user already selected an installed skill and only wants to run it, or the request is unrelated to skill discovery or installation.
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.