Skills
96,770 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Builds AI agents on Cloudflare using the Agents SDK with state management, real-time WebSockets, scheduled tasks, tool integration, and chat capabilities. Generates production-ready agent code deployed to Workers. Use when: user wants to "build an agent", "AI agent", "chat agent", "stateful agent", mentions "Agents SDK", needs "real-time AI", "WebSocket AI", or asks about agent "state management", "scheduled tasks", or "tool calling". Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
Provides LangChain4j patterns for implementing MCP (Model Context Protocol) servers, creating Java AI tools, exposing tool calling capabilities, and integrating MCP clients with AI services. Use when building a Java MCP server, implementing tool calling in Java, connecting LangChain4j to external MCP servers, or securing tool exposure for agent workflows.
Ultra-compressed communication mode. Cuts token usage by about 75% by speaking like a caveman while preserving full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when the user says “caveman mode,” “talk like caveman,” “use caveman,” “less tokens,” “be brief,” or invokes /caveman. Also auto-triggers when token efficiency is requested.
Create and maintain gjkim_instruction.md, the root document for a loop-engineering effort. The document holds only the minimum requirements and confirmed decisions, and deliberately leaves everything else open so that later iterations are not locked into early guesses. Use whenever the user mentions gjkim_instruction.md, a root document or root doc for a loop, loop engineering, starting a long-running agent loop on a goal, or asks to record a requirement or a confirmed decision for such an effort. Also use at the start of a loop iteration to check what is already decided versus still open.
Delegate a coding task to the Pi coding agent CLI (`pi`) as a background implementer, then review its diff and land it yourself. Use this whenever the user wants to delegate implementation work to Pi - phrasings like "have Pi implement X", "delegate this to pi", "run it through Pi", or "use pi to implement/fix/refactor" - or wants to run a queue of coding tasks through Pi 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.
Execute a plan or concrete work prompt end-to-end. Use when implementing from a plan document, a spec path, or a clear build request; use ce-debug for open-ended bugs. Use when an outer orchestrator needs implementation and local verification only, without the shipping tail.
Write, run, and analyze structured test suites for Agentforce agents. TRIGGER when: user writes or modifies test spec YAML (AiEvaluationDefinition); runs sf agent test create, run, run-eval, or results commands; asks about test coverage strategy, metric selection, or custom evaluations; interprets test results or diagnoses test failures; asks about batch testing, regression suites, or CI/CD test integration. DO NOT TRIGGER when: user creates, modifies, previews, or debugs .agent files (use developing-agentforce); deploys or publishes agents; writes Agent Script code; uses sf agent preview for development iteration; analyzes production session traces (use observing-agentforce).
Break a plan, spec, or current conversation into a set of tracer-bullet tickets. Each ticket declares blocking edges, using text edges in one file per ticket locally and native blocking links in a real tracker. Publish them to the configured tracker.
Scaffolds new agent skills for the dotnet/skills repository. Use when creating a new skill, generating SKILL.md files, writing a skill description that the runtime will actually route to, or setting up skill directory structures. Handles frontmatter generation, section templates, and validation guidance. Do not use for fixing a skill that already fails its evaluation (use improve-skill-quality) or for writing eval.yaml (use create-skill-test).
Reconstruct your recent working context from your own chat history, live state, and the shared record (user reports, prior fixes, incidents), then hand back a tight current-state brief. Use for “recall my work on X”, “catch me up”, “what have I been working on”, “where did I leave off”, before starting or resuming work.
Orchestrate Xcode build optimization by benchmarking first, running the specialist analysis skills, prioritizing findings, requesting explicit approval, delegating approved fixes to xcode-build-fixer, and re-benchmarking after changes. Use when a developer wants an end-to-end build optimization workflow, asks to speed up Xcode builds, wants a full build audit, or needs a recommend-first optimization pass covering compilation, project settings, and packages.
One-time pipeline configurator. Inspects the repo (default branch, validation scripts, labels), asks a few questions, writes .ai/agentic.config.json — the file every other skill reads — installs the tracker descriptor, and generates missing project docs (SDLC.md, CODE_REVIEW.md, BACKWARD_COMPATIBILITY.md, AGENTS.md starter). Re-run when the toolchain or label taxonomy changes. Verifies cross-skill coverage and prints the install command for missing skills.
Analyze production Agentforce agent behavior using session traces and Data Cloud. TRIGGER when: user queries STDM session data or Data Cloud trace records; investigates production agent failures, regressions, or performance issues; asks about session traces, conversation logs, or agent metrics; wants to reproduce a reported production issue in preview; runs findSessions or trace analysis queries. DO NOT TRIGGER when: user creates, modifies, or debugs .agent files during development (use developing-agentforce); writes or runs test specs (use testing-agentforce); uses sf agent preview for local development iteration; deploys or publishes agents.
Drive an open PR to merge-ready from its number — merges the latest base, then loops review-autofix (om-auto-review-pr), built-in CI stabilization, and UI verification (om-auto-qa-pr) until approvable, green, and QA-evidenced. A --ci-only mode drives just CI green on a PR or a plain branch. Files follow-up issues for nits, normalizes labels, hands off to om-approve-merge-pr — never merges itself. Use for "get PR 123 merge-ready".
Documentation reference for writing Python code using the browser-use open-source library. Use this skill whenever the user needs help with Agent, Browser, or Tools configuration, is writing code that imports from browser_use, asks about @sandbox deployment, supported LLM models, Actor API, custom tools, lifecycle hooks, MCP server setup, or monitoring/observability with Laminar or OpenLIT. Also trigger for questions about browser-use installation, prompting strategies, or sensitive data handling. Do NOT use this for Cloud API/SDK usage or pricing — use the cloud skill instead. Do NOT use this for directly automating a browser via CLI commands — use the browser-use skill instead.
Guides technology selection and implementation of AI and ML features in .NET 8+ applications using ML.NET, Microsoft.Extensions.AI (MEAI), Microsoft Agent Framework (MAF), GitHub Copilot SDK, ONNX Runtime, and OllamaSharp. Covers the full spectrum from classic ML through modern LLM orchestration to local inference. Use when adding classification, regression, clustering, anomaly detection, recommendation, LLM integration (text generation, summarization, reasoning), RAG pipelines with vector search, agentic workflows with tool calling, Copilot extensions, or custom model inference via ONNX Runtime to a .NET project. DO NOT USE FOR projects targeting .NET Framework (requires .NET 8+), the task is pure data engineering or ETL with no ML/AI component, or the project needs a custom deep learning training loop (use Python with PyTorch/TensorFlow, then export to ONNX for .NET inference).
AI-native tutor and onboarding workflow for the four independent Claude certification tracks in AI Engineering from Scratch. Use when a learner wants to choose a Claude certification, prepare for CCAO-F, CCDV-F, CCAR-F, or CCAR-P, resume a certification path, learn the next lesson interactively, run and verify practical labs, build scored artifacts, take a diagnostic or mock exam, or remediate weak exam domains from GitHub with Claude Code, Codex, ChatGPT, Cursor, or another agent
Provides a structured 8-phase workflow for resolving GitHub issues in Claude Code. Covers fetching issue details, analyzing requirements, implementing solutions, verifying correctness, performing code review, committing changes, and creating pull requests. Use when user asks to resolve, implement, work on, fix, or close a GitHub issue, or references an issue URL or number for implementation.
Implement an existing spec and ship a verified, reviewed, ready PR. Resolves the spec by path, name, issue, or spec-PR number (clean stop with candidates when not found). A spec PR stays design-only — implementation ships on its own PR referencing it. Delegates to om-auto-create-pr (om-auto-continue-pr for an existing implementation PR), then runs the review loop and UI verification with screenshots. Use for "implement the spec X", "build spec from issue 123".
Take a request all the way to done, hands-off, through the right Compound Engineering skills. A code change ends as an open pull request, pushed without stopping. Use only when the user explicitly asks for autonomous end-to-end work or invokes lfg directly. Use ce-plan, ce-work, ce-debug, or ce-commit-push-pr for work the user reviews step by step.
Continue working on an OpenSpec change by creating the next artifact. Use when the user wants to progress their change, create the next artifact, or continue their workflow. Also use when the user says "openspec continue" or "opsx continue".
Provides comprehensive memory file management capabilities including auditing, quality assessment, and targeted improvements for files such as CLAUDE.md. Use when user asks to check, audit, update, improve, fix, maintain, or validate project memory files. Also triggers for "project memory optimization", "CLAUDE.md quality check", "documentation review", or when a project memory file needs to be created from scratch. This skill scans memory files, evaluates quality against standardized criteria, outputs detailed quality reports with scores and recommendations, then makes targeted updates with user approval.
Use this skill when the user asks to add, embed, integrate, configure, style, or remove an agent, chatbot, chat widget, conversation client, or AI assistant in a UI Bundle project. Trigger when: the project contains a uiBundles/*/src/ directory and the task involves adding or modifying a chat widget, chatbot, or conversational AI; files under uiBundles/*/src/ import AgentforceConversationClient; or the user asks to add any chat or agent functionality to a page. Do NOT trigger when: the user wants to create a custom agent, chatbot, or chat widget component from scratch; or the project has no uiBundles directory
Guidance on how to use Claude Code effectively — covering context management, verification strategies, the explore-plan-implement workflow, prompting techniques, session management, parallel sessions, and common failure patterns. Use this skill whenever the user asks how to get the most out of Claude Code, how to write better prompts, how to manage context, when to use plan mode, how to automate tasks, or when they describe a frustrating pattern like Claude repeating mistakes or losing track of instructions.
Guide for creating, structuring, and improving Claude skills (SKILL.md). Use when building a new skill, reviewing an existing skill, writing SKILL.md frontmatter, defining trigger conditions, troubleshooting skill problems (not triggering, over-triggering, instructions not followed), or planning skill distribution. When working on any skill in this repository: also load the cc-best-practices skill, and always update both CLAUDE.md and README.md skill tables after any skill change. Do NOT use for general Claude Code configuration or hook setup
Guides the design and structuring of workflow-based Claude Code skills with multi-step phases, decision trees, subagent delegation, and progressive disclosure. Use when creating skills that involve sequential pipelines, routing patterns, safety gates, task tracking, phased execution, or any multi-step workflow. Also applies when reviewing or refactoring existing workflow skills for quality.
Provides Gemini CLI delegation workflows for large-context analysis and complex reasoning using Gemini 3.0 Flash and Gemini 3.0 Pro models, including English prompt formulation, execution flags, and safe result handling. Use when the user explicitly asks to use Gemini for tasks such as broad codebase analysis, fast iterations with Gemini 3 Flash, or deep architectural reasoning with Gemini 3 Pro. Triggers on "use gemini", "delegate to gemini", "run gemini cli", "ask gemini", "use gemini for this task", "use gemini 3 flash", "use gemini 3 pro".
Scaffolds eval.yaml evaluation specs for skills, custom agents, and redistributable gh-aw workflow packages in the dotnet/skills repository. Use when creating skill or workflow-package tests, writing evaluation stimuli, defining graders and rubrics, sizing an eval for statistical power, or setting up test fixture files. Handles the Vally eval.yaml schema, fixture organization, and overfitting avoidance. Do not use for running or debugging existing evals (use improve-skill-quality) nor for skills authoring (use create-skill).
Provides GitHub Copilot CLI task delegation in non-interactive mode with multi-model support (Claude, GPT, Gemini), permission controls, output sharing, and session resume. Use when users ask to hand work to Copilot, compare models, or run Copilot programmatically from Claude Code.
Start a new OpenSpec change using the experimental artifact workflow. Use when the user wants to create a new feature, fix, or modification with a structured step-by-step approach. Also use when the user says "openspec new change" or "opsx new"
Autonomous skill generator that learns new technologies from the web. Use when, users want to learn about a new library/framework/tool, need to create a skill for an unfamiliar technology, want to research and document a technology's usage patterns, or invoke with `/learn <topic>`. This skill uses web search and browser tools to discover, extract, and synthesize documentation into a reusable skill.
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". Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
Use when the user asks for a code review by a fleet of specialized reviewer agents, wants multiple independent reviewer perspectives, or asks to run reviewers in single-pass or iterative fix-until-clean mode. Launches focused subagents for correctness, security, architecture, conventions, simplicity, UX, reliability, telemetry, testing, compatibility, and documentation review.
Autonomous iterative experimentation loop for any programming task. Guides the user through defining goals, measurable metrics, and scope constraints, then runs an autonomous loop of code changes, testing, measuring, and keeping/discarding results. Inspired by Karpathy's autoresearch. USE FOR: autonomous improvement, iterative optimization, experiment loop, auto research, performance tuning, automated experimentation, hill climbing, try things automatically, optimize code, run experiments, autonomous coding loop. DO NOT USE FOR: one-shot tasks, simple bug fixes, code review, or tasks without a measurable metric.
Provides Codex CLI delegation workflows for complex code generation and development tasks using OpenAI's GPT-5.3-codex models, including English prompt formulation, execution flags, sandbox modes, and safe result handling. Use when the user explicitly asks to use Codex for complex programming tasks such as code generation, refactoring, or architectural analysis. Triggers on "use codex", "delegate to codex", "run codex cli", "ask codex", "codex exec", "codex review".
Knowledge and governance closeout: reconcile project docs, rule files (CLAUDE.md/AGENTS.md), authorized agent memory, and workspace residue with what the code and runtime actually do, so the next session or the next person starts from one current answer. Trigger when the user names "neat-freak", "洁癖", or "/neat" — and also on clear knowledge-closeout intent without the name: syncing or tidying project docs/rules/memory after development ("把文档和记忆整理一下", "收尾时把文档同步掉", "docs 和代码对不上了"), stale or conflicting CLAUDE.md/memory, a clean handoff to a teammate or a fresh session, or auditing whether workspace rules are actually followed. Do not trigger for pure coding/refactoring/debugging tasks, tidying data or prose (JSON, 周报, changelog announcements), or a bare "整理" with no project-knowledge context.
Generate and evaluate grounded ideas. Use when the user wants ideas, improvements, or surprising directions before choosing one to develop. Not for refining an idea they already have (ce-brainstorm) or judging one already on the table (ce-pov).
Delegate a coding task to the Qoder CLI (`qodercli`) as a background implementer, then review its diff and land it yourself. Use this whenever the user asks to have Qoder implement, fix, refactor, or run a queue of coding tasks while the orchestrator remains the reviewer. DO NOT USE for tasks small enough to do inline, or when the user wants code written directly without delegation.
Systematic architectural thinking for irreplaceable human capabilities - domain modeling, systems thinking, constraint navigation, and AI-aware problem decomposition. Use proactively when detecting architectural decisions, system design discussions, or multi-component planning.
Use when users want to run a local coding agent on KIE's chat models (api.kie.ai), ask which KIE models an agent can use, or ask how to configure its base URL, API key, provider table, or model name. Currently covers Codex CLI, Claude Code, and Grok Build. Also use this skill when the model's wire protocol differs from the agent's and a local translation proxy is needed.
Iteratively inspect traces, interview the user, and create LangSmith online evaluators one at a time. Use specifically for creating online evaluators for use within LangSmith -- use "eval-engineering" for Harbor-style online evaluations.
Provides autonomous project pattern learning by analyzing the codebase to discover development conventions, architectural patterns, and coding standards, then generates project rule files in .claude/rules/. Use when the user asks to "learn from project", "extract project rules", "analyze codebase conventions", "discover project patterns", or wants to auto-generate Claude Code rules for the current project.
Vision-driven iOS device automation using Midscene CLI. Operates entirely from screenshots — no DOM or accessibility labels required. Can interact with all visible elements on screen regardless of technology stack. Control iOS devices with natural language commands via WebDriverAgent. Triggers: ios, iphone, ipad, ios app, tap on iphone, swipe, mobile app ios, ios device, ios testing, iphone automation, ipad automation, ios screen, ios navigate, test ios app, verify on iphone, QA on ipad, check the app on ios, test on ios device, see if the app works on iphone, end-to-end test on ios, visual verification on ios Powered by Midscene.js (https://midscenejs.com)
Unified TUI and CLI to index and search your local coding agent session history across 30+ providers (Codex, Claude Code, Gemini, Cursor, Aider, Copilot, and more) into one searchable timeline. Hybrid search plus robot/JSON output for agents
Optimizes, improves, and debugs LLM prompts using production trace data, evaluations, and annotations. Extracts prompts from spans, gathers performance signal, and runs a data-driven optimization loop using the ax CLI. Use when the user mentions optimize prompt, improve prompt, make AI respond better, improve output quality, prompt engineering, prompt tuning, or system prompt improvement.
Review skills in any project using a dual-axis method: (1) deterministic code-based checks (structure, scripts, tests, execution safety) and (2) LLM deep review findings. Use when you need reproducible quality scoring for `skills/*/SKILL.md`, want to gate merges with a score threshold (for example 90+), or need concrete improvement items for low-scoring skills. Works across projects via --project-root
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.