Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Review recent repository changes and decide whether AGENTS.md or other project-level documentation needs a high-level update. Use when finishing a feature, fix, refactor, or architectural change and you need to preserve repo-shaping guidance such as new patterns, constraints, workflows, validation rules, or onboarding-relevant gotchas without adding low-level implementation detail.
Guidance for Microsoft Security Copilot - the generative-AI security platform that helps analysts investigate, hunt, summarise, and respond using natural language, plugins, promptbooks, and embedded experiences. Covers SCU provisioning, plugins, promptbooks, governance, and the standalone vs embedded experience choice. WHEN: Microsoft Security Copilot, AI for SOC, security copilot units SCU, promptbooks, Copilot plugins, natural language investigation, summarise incident with AI, Copilot for Security setup, how do I use AI in my SOC, explain a KQL query with AI, summarise an alert for a stakeholder. DO NOT USE when the goal is configuring autonomous triage or remediation agents (use security-copilot-agents).
Use when starting work on a new or unfamiliar project, when encountering unexpected patterns, when user corrects your assumptions, or when explicitly invoked via /learn - auto-discovers and remembers project context through structured codebase analysis
Use when you have a rough product idea and want a complete PRD without sitting through an interactive grilling. Claude walks the full decision tree (edge cases, modules, schema, testing, security), self-answers with software-engineering best practices, streams the Q&A live so you can override, and writes the PRD locally with an option to push as a GitHub issue.
Design, create, validate, review, and package a Claude Code skill (SKILL.md plus references/scripts/assets). Use when asked to 「スキル作って」「スキル設計」「SKILL.md 作成」「スキルレビュー」, when a new reusable workflow or domain knowledge should become a skill, when an existing skill's description fails to trigger or its SKILL.md has grown past progressive-disclosure limits, or when a skill needs a sanitization/security review before distribution. Do NOT use for: running or invoking an existing skill, or editing CLAUDE.md / rules (those are not skills).
Find and evaluate Claude skills for specific use cases using semantic search, Anthropic best practices assessment, and fitness scoring. Use when the user asks to find skills for a particular task (e.g., "find me a skill for pitch decks"), not for generic "show all skills" requests.
Guide to analyze a current work session and propose improvements to skills. Use (1) automatically after working with a skill to capture learnings, (2) when the user suggests improvements, corrections, or additions during a skill-related session, or (3) when the user manually invokes `self-improve`.
A skill that assists with creating and editing TAKT's tasks.yaml (task metadata) and task directories (.takt/tasks/{slug}/order.md). It generates YAML entries conforming to the TaskRecord schema, creates order.md task specifications, and validates status transition rules. Uses the task schema definitions and documentation in references/takt as reference material. Triggers: 「タスクを追加」「tasks.yamlを編集」「taktタスクを作成」「タスク仕様書を書く」「order.mdを作成」「takt task」「タスクを定義」「pendingタスクを追加」「GitHub Issueからタスク作成」
Implements work already planned—a plan.md from tlc-plan or a small ticket—by freezing it into checks with proofs, building, and proving every check with a script and an independent Verifier before declaring done. Use when the user says “implement this plan”, “build this ticket”, “implement this spec”, “write the checks”, “verify work”, or “tlc-implement”. Do NOT use when nobody has decided what to build, or to plan a feature that needs one (that is tlc-plan).
Turns a request, ticket, PRD, design or thread into one plan a human approves and a builder can act on without the conversation: asks every gray area until no product decision is left to the builder, then writes the problem, EARS criteria with concrete values, the path and the data it touches (drawn in mermaid), the interface, and only the decisions that are hard to reverse. Captures that come with the request bind. Use when the user says “plan this”, “plan feature”, “specify feature”, “write the plan”, “cut this PRD into tasks”, or “tlc-plan”. Do NOT use to implement (that is tlc-implement), for a standalone design document, or when nobody has decided what to build.
Split a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker — as text edges in one file per ticket locally, or as native blocking links on a real tracker.
Store and retrieve conversation context using Twilio Conversation Memory. Covers Memory Store provisioning, profile management, traits, observations, conversation summaries, and semantic Recall. Use this skill to give AI agents or human agents persistent memory of conversations across sessions and channels.
Use when starting any conversation, receiving a new task, or when uncertain which skill applies - establishes how to find and use all 64 toolkit skills, requiring Skill tool invocation before ANY response including clarifying questions
Use when the user needs to build AI agents — tool use patterns, memory management, planning strategies, multi-agent coordination, evaluation, and safety guardrails. Triggers: user says "agent", "build an agent", "tool use", "agent loop", "multi-agent", "memory management", "guardrails", "agent evaluation".
Generate hierarchical AGENTS.md structures for codebases. Use when user asks to create AGENTS.md files, analyze codebase for AI agent documentation, set up AI-friendly project documentation, or generate context files for AI coding assistants. Triggers on "create AGENTS.md", "generate agents", "analyze codebase for AI", "AI documentation setup", "hierarchical agents".
[Hyper] Use this skill when optimizing an existing Codex skill through baseline-first experiments, binary evals, Guard checks, and one-mutation-at-a-time iteration. Use for skill autoresearch, measured trigger/workflow improvement, self-optimizing a skill, benchmarking skill changes, or resuming skill experiment artifacts. If invoked without a target, ask for the target skill and eval intent before creating artifacts or mutating files. Do not use for one-off skill creation/refactor, generic docs polish, app QA, commit-only, or push-only requests.
Benchmark Blender modeling agents, skills, prompts, scripts, or MCP tools with paired isolated runs. Use for baseline-versus-plugin comparisons, regression suites, skill forward-testing, MCP usefulness evaluation, score calibration, or claims that a Blender workflow improves mesh, visual, structural, animation, cost, or completion quality.
Use when a problem or idea has been talked through — in Superpowers' brainstorming or any other conversation — and the user wants it tracked as an intent issue instead of built now: "file this", "open an issue for this", "track it first", or picking track-first when superpowers-issue-bridge asks. Not while the problem is still being explored.
Copilot left 14 review comments on your PR — half are nits. Hours of fix → reply → resolve → re-request, and each round lands MORE comments. This skill runs loop engineering: auto-triggers Copilot Code Review via GraphQL (no @copilot mention), triages every open thread (Copilot, humans, advanced-security) with a fix / decline / escalate rubric, dispatches parallel fix sub-agents that obey the repo build/test/lint conventions, commits per iteration, replies+resolves citing the pushed SHA, then re-triggers until HEAD is reviewed with zero threads awaiting the agent's reply (remaining open threads are explicit hand-offs to the human — escalated declines, design tradeoffs). You merge a clean PR; the bot runs it. Trigger phrases: "address copilot comments", "run a copilot review loop", "fix this PR", "iterate on copilot feedback". Repo-agnostic, gh CLI + PowerShell. Full autopilot needs repo Triage/Write; external PR authors get single-iteration mode plus manual re-trigger (UI 🔄 or substantive-commit push).
Use this skill when the user asks to create, scaffold, update, or review a MoviePilot agent skill. This includes adding a new built-in skill under the repository `skills/` directory, editing an existing built-in skill, writing `SKILL.md` frontmatter and workflow instructions, choosing `allowed-tools`, adding helper scripts when needed, and bumping the built-in skill `version` so changes can sync into `config/agent/skills`.
Doubaoya (doubaoya / 本鸭) skill reconciliation updater: reconciles the locally installed set of doubaoya skills to the current full set in the official repository — archiving ones removed upstream, installing newly added ones, refreshing outdated ones, then self-checking. Packages are identified by content hash, so other people's and user-modified ones are left untouched down to a single character; removed ones are moved to an archive directory rather than deleted. Reconciles the installed doubaoya skills to the upstream set, then self-checks. Triggered by: /dby-update, 更新本鸭, 更新都爆鸭, 升级 doubaoya skill, 检查本鸭更新, 把本鸭更新到最新版.
Guidelines for de-jargoning Chinese output. Activates automatically when the model replies in Chinese, keeping output concise, direct and natural, and avoiding AI-flavored expressions such as consulting jargon, internet slang, fake urgency and emotional manipulation. Trigger phrases: 「讲人话」「去油」「去黑话」「去 AI 味」「别那么 GPT」「正常说话」. Even without a trigger phrase, the rules of this skill should apply whenever the model is generating a Chinese reply. Also applies when the user says 「这段话太 GPT 了」「帮我改成人话」「输出太油了」「说话别绕」.
Dispatches one subagent per independent domain to parallelize investigation/fixes. Use when you have 2+ unrelated failures (e.g., separate failing test files, subsystems, bugs) with no shared state or ordering dependencies.
Use when a task has multiple independent subtasks that can be executed concurrently by separate agents. Triggers when decomposed work has 2+ subtasks with no data dependencies, when subtasks operate on different files or codebase sections, when serial execution time would be significantly longer than parallel, or when independent analyses or deliverables need concurrent generation.
Use when you have an approved implementation plan document and need to execute it step by step. Triggers on /execute command, when transitioning from planning with an approved plan, or when resuming execution of a partially completed plan. Provides batch-based execution with TDD, checkpoint reviews, and verification gates.
Structures the human review experience for factory-mode builds with dashboard-style audit trail summaries (slice counts, rework rates, gate failure distribution), PR digests (per merged PR with gate results), retrospective synthesis (recurring rework causes, pair effectiveness, domain hotspots), quality trend tracking (session-over-session comparison), escalation review (rework history with resolve/override/send-back options), and autonomy tuning interface. Use during Phase 3 human review of factory-mode builds or when assessing autonomous build quality. Triggers on: "review the build", "factory review", "audit trail", "quality trends", "escalation review", "PR digest", "autonomy tuning", "rework rate". NOT for: running the pipeline (use pipeline), individual code reviews (use code-review).
Invoke when the user wants to start a new game project, set up project scaffolding, onboard into an existing codebase, or initialize GameForge session state. Triggers on: "start a game", "new game project", "set up project", "project onboarding", "initialize project". Do NOT invoke for brainstorming ideas (use game-brainstorm) or prototyping mechanics (use game-prototype). Part of the AlterLab GameForge collection.
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.