Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Set up and optimize repositories for AI coding agents. Creates minimal AGENTS.md, CLAUDE.md symlink, docs/REQUIREMENTS.md, docs/BUSINESS-RULES.md, feedback loops, and deterministic enforcement (Claude Code hooks, OpenCode plugins). Use when user wants to make a repo AI-friendly, set up AGENTS.md/CLAUDE.md, document requirements/business rules for AI, add pre-commit hooks for AI workflows, or optimize codebase structure for coding agents.
Improves an existing application using its code, product contract, authorized work history and competitor evidence. Use for 取长补短、取其精华去其糟粕、参考竞品优化应用 or evolving an app from successful and failed work. Carries authorized changes through implementation and user-result verification, including REA and reverse Skill composition; use skill-evolver for Skills.
Use for "automate me", "create/update/refresh my -mode skill", "turn/capture my preferences or working style into a skill", or wanting agents to follow how the user works. Drafts or revises a personal -mode skill via create-skill + unslop, optionally pulling fresh evidence from recent transcripts
Extract core methodologies from books and generate the corresponding skill. Use this skill when the user says "extract the methodology from 《XXX》", "generate a skill for 《XXX》", "turn this book's methodology into a skill", or has a similar request. Suitable for methodology extraction and skill generation from books of various kinds, including business, management, personal growth, and technical books.
Design-to-plan host: thin wrap of Superpowers brainstorming, then hard handoff to solve-workflow Make a Plan. Use for new features and fuzzy design before coding. Triggers — 「头脑风暴」「设计头脑风暴」「brainstorm」「brainstorm workflow」「先设计再计划」「头脑风暴工作流」. Do NOT use for bug root-cause PDCA (solve-workflow / opsx-solve-workflow) or when you only need a one-line edit.
Methodology for Playwright Browser MCP and xterm.js terminal interaction. Use when you need to operate terminals embedded in web pages through a browser, such as pseudo-terminals in CTF ranges, Cloud Shell, or terminals in online IDEs. Covers terminal content reading (5 methods), command execution, output capture, and screenshot fallback strategies. Use this skill whenever the target page contains any form of web terminal (xterm.js/hterm/jQuery Terminal)
Builds remote MCP (Model Context Protocol) servers on Cloudflare Workers with tools, OAuth authentication, and production deployment. Generates server code, configures auth providers, and deploys to Workers. Use when: user wants to "build MCP server", "create MCP tools", "remote MCP", "deploy MCP", add "OAuth to MCP", or mentions Model Context Protocol on Cloudflare. Also triggers on "MCP authentication" or "MCP deployment".
OpenClaw system diagnostics and performance analysis tool. Analyzes agent reasoning time, Token usage, tool call statistics, Run timelines, and Gateway restart history. Supports multiple modes: batch analysis (default), real-time tracking (-f), summary statistics (-s), advanced diagnostics (--advanced). Supports multi-Agent filtering. Use cases: triggered when users ask about OpenClaw running status, performance bottlenecks, reasoning latency, Token consumption, tool execution statistics, error troubleshooting, or agent activity analysis. Trigger words: diagnostics, performance analysis, reasoning time, token statistics, running status, agent analysis, tool call statistics, Run details, Gateway restart. Command trigger: /diag — runs diagnostics directly, supports parameter passthrough.
Analyzes professions/jobs for Claude Code automation opportunities. Triggers: 'how can Claude help me', 'what can Claude do for', 'I'm a [profession]', 'help me as a [job]', 'I work as', describing their work + asking about Claude. Use whenever user mentions their profession/role and wants to discover what Claude can automate.
Configure CLAUDE.md project memory files for persistent context, coding standards, architecture decisions, and team conventions. Reference for the 4-tier memory hierarchy, cross-platform AGENTS.md compatibility, and quick-add commands.
Manage Claude Code sessions — search by topic or ID, browse recent sessions with full metadata (tags, summaries, titles, cost), view title/nickname history timelines, resume in Ghostty tabs, spawn interactive or headless sessions, monitor live sessions, and bootstrap new setups. Triggers on resume session, find session, list sessions, recent sessions, spawn session, session history, what was I working on, open in ghostty, title history, session nicknames.
Implementation subagent dispatch patterns. Use when a task's design is settled and the file just needs to be made, or when independent implementation work can run in parallel. Covers eligibility, dispatch, shared-tree patch snapshots, one combined review per completed batch, and serial integration.
An intelligent multi-store memory system with human-like encoding, consolidation, decay, and recall. Use when setting up agent memory, configuring remember/forget triggers, enabling sleep-time reflection, building knowledge graphs, or adding audit trails. Replaces basic flat-file memory with a cognitive architecture featuring episodic, semantic, procedural, and core memory stores. Supports multi-agent systems with shared-read, gated-write access. Includes philosophical meta-reflection that deepens understanding over time. Covers MEMORY.md, episode logging, entity graphs, decay scoring, reflection cycles, evolution tracking, and system-wide audits.
Control-plane workflow for coordinating multi-agent, multi-session project work from a single Codex, GitHub Copilot, or agent-app control session. Use this skill whenever the user asks to orchestrate agents, create or steer worker sessions, run a workflow-like effort, fan out audits/research/migrations, coordinate parallel implementation streams, monitor other project sessions, or compare this control-session pattern to Claude Code dynamic workflows. This skill is especially relevant when the current session can spawn persistent project sessions and those sessions can spawn their own subagents, creating a two-level orchestration hierarchy.
Find the mistakes agents keep repeating in this repo and make each one impossible. Try architecture first, then types, then a lint whose error names the fix, then a test, and write docs last. Prove each check fails on a real past mistake. Repeat this each time the operator corrects you. Use for /correct.
Creates production-ready MCP servers using FastMCP v2. Use when building MCP servers, optimizing tool descriptions for context efficiency, implementing progressive disclosure for multiple capabilities, or packaging servers for distribution.
Cuihuo Thinking v2, the general thinking entry point and topic hub of the 边淬火 family. It takes unformed input such as "I'm thinking about", "I have a question", "this needs to be thought through", "I'm stuck", "I can't make sense of it" or "I'm thinking about that thing from last time", runs an old-topic check, identity check and language/material/function audit, routes to Investigation, Containment, Dissolution, Cuihuo Diagnosis, Cuihuo Questioning, Cuihuo Verification, Cuihuo Learning, Cuihuo Brainstorm, Cuihuo Logic Chain, Cuihuo Judgment Formation, Cuihuo Decision or GTD, and updates the topic once results come back. It replaces the old "Thinking" skill; do not intercept when the user has already explicitly named a downstream skill.
Universal deep research agent team. 13-agent pipeline for rigorous academic research on any topic. 8 modes: full research, quick brief, paper review, lit-review, fact-check, three-way literature scan, Socratic guided research dialogue, and systematic review with optional meta-analysis. Covers research question formulation, Socratic mentoring, methodology design, systematic literature search, source verification, cross-source synthesis, risk of bias assessment, meta-analysis, APA 7.0 report compilation, editorial review, devil's advocate challenges, ethics review, and post-research literature monitoring. Triggers on: research, deep research, literature review, systematic review, meta-analysis, PRISMA, evidence synthesis, fact-check, WHY HOW WHAT papers, 3W literature scan, guide my research, help me think through, 研究, 深度研究, 文獻回顧, 文獻探討, 系統性回顧, 後設分析, 事實查核, 三段式文獻掃描, 引導我的研究, 幫我釐清, 幫我想想, 我不確定要研究什麼, 研究方向, 研究主題.
Open a Diffo code review of the current git changeset, or of a GitHub pull request the user links, in the reviewer's browser and drive the feedback loop from this session, using the diffo CLI. Use when the user wants to review your changes, asks to open a code review, hands you a PR to review with them, or when you finish a large or multi-file change worth a human read before it lands.
Use this skill when you need semantic codebase context before planning, explaining, or changing project code. It runs a bundled local Fast Context script that calls Windsurf Devstral directly, avoiding the Fast Context MCP server while preserving natural-language repository search, recursive context gathering, and definition/signature completeness checks.
This skill should be used when taking a single software feature from intent to shipped as a solo developer — goal-first, TDD, deterministic verification, evidence only where it earns its keep, and human judgment at the two moments that matter (goal approval, merge). Trigger when the user says "let's build feature X", "ship this feature", "run this through the factory", "write a goal contract", "feature-factory", or wants a disciplined intent→merge loop that resists process bloat. NOT for whole-product planning, multi-feature roadmaps, or autonomous multi-agent swarms.
Design an auditable playbook when no narrower one fits: a large migration, an ambitious multi-part change, or work a human reviews after stepping away. Scale rigor to the task, run a hypothesis loop, and log decisions. Use for /figure-it-out or "figure it out".
Runs a multi-step procedure as a Python DAG, so ordering, branching and retries are enforced by the runner rather than described in prose a model can generate past. Use for "run these steps in order and retry the flaky one until the check passes", "build a pipeline that fetches, validates, then skips the upload when nothing changed", "make sure these steps cannot be skipped", "resume from where it broke instead of redoing the expensive early stages", "run these independent calls at once and merge the results", or any procedure of 3+ steps with branches, input contracts, or side effects that must not block the critical path. Primitives are depends_on, when=, validate=, retry_until=, detached= and journal_path=. Not for a single sequential call, for steps needing reasoning between them that no predicate captures, or for async and distributed work. To audit whether one verification check can actually go red, use gating. To fan work out across many subagents, use a dynamic workflow.
Manus-level AI browser agent that drives Dhruv's REAL logged-in Chrome (any profile — AIV, Onsite, etc.) via natural-language commands. Persistent, human-paced, un-blockable (real session), LLM-driven so it survives obfuscated selectors. Use when Dhruv says "control my AIV / control my <profile>", or asks to browse, search, apply, engage, or automate anything in his real browser.
Pre-write (docs or code) git/native worktree isolation with worktree-gate (always|never|ask; unset≡ask): detect reuse (incl. submodule guard), recommend suitability, create (native then ignored .worktrees/), setup+baseline, or decline/lean with 留痕; cleanup via feature-branch-closeout. Triggers — 「git worktree」「工作区隔离」「隔离工作区」「执行前隔离」「worktree 门控」 / git worktree discipline. Do NOT use as Superpowers using-git-worktrees alias or for closeout destroy.
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.