Skills
97,725 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Execute implementation tasks from the current plan. Works through tasks sequentially, marks completion, and preserves progress for continuation across sessions. Use when user says "implement", "start coding", "execute plan", or "continue implementation".
Use when the user says "autopilot", "handle it yourself", or "just run it to the end", or delegates the agent to investigate, implement, and verify a broad task. For requests that are only planning, explanation, or already have a clear narrow scope, handle them within the scope the user specifies.
Set up and run an autonomous experiment loop for any optimization target. Gathers what to optimize, then starts the loop immediately. Use when asked to "run autoresearch", "optimize X in a loop", "set up autoresearch for X", or "start experiments".
Stop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP, Claude Code, Claude Cowork, Claude in Excel, Claude in PowerPoint, Claude in Chrome, the Claude API, or prompting technique; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together, even when another skill has already answered the question. Only recommend on a strong match; never invent Academy content.
Manages permanent memory storage for decisions, blockers, context, preferences, and procedures. Use when user says "remember", "save this decision", "what did we decide", "recall", "search memories", "any blockers", or when making important architectural decisions. Provides SDAM compensation through external memory.
After giving a substantive answer or draft the user may act on — such as advice, recommendations, drafted artifacts, estimates, projections, analysis or interpretation of data, factual claims they may rely on, or a multi-step argument — invoke this skill before finalizing your reply. If applicable, append 2–3 short follow-up questions tied to specifics in your response to help the user check key facts, examine reasoning or assumptions, and spot missing context. Do this at most once per conversation. Skip it for trivial how-tos or simple lookups, purely educational explanations, formatting/conversion/assembly of user-provided content, code the user will run, creative writing, casual chat, or requests to double-check, cite, or review. See the skill file for boundaries and exact output format.
Automate construction data processing using LLM (ChatGPT, Claude, LLaMA). Generate Python/Pandas scripts, extract data from documents, and create automated pipelines without deep programming knowledge.
Transforms vague or unreliable prompts into structured, constraint-aware prompts with explicit roles, task decomposition, output formats, and quality checks. Use when prompts produce inconsistent outputs, need explicit structure and constraints, require safety guardrails, involve multi-step reasoning that needs decomposition, need domain expertise encoding, or when user mentions improving prompts, prompt templates, structured prompts, prompt optimization, reliable AI outputs, or prompt patterns.
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Claude ('No, that's wrong...', 'Actually...'), (3) User requests a capability that doesn't exist, (4) An external API or tool fails, (5) Claude realizes its knowledge is outdated or incorrect, (6) A better approach is discovered for a recurring task. Also review learnings before major tasks.
Triage Roblox game slices and bug reports through a small state machine so only fully-specified, playtestable work gets built. Use when the user wants to review incoming slices or bugs, decide what to build next, prepare a slice for building, mark work ready, or manage the roadmap/issue workflow. Keeps Claude from picking up vague, half-specified work and building the wrong thing.
Teaches AI agents to recognize and avoid security threats during normal activity. Covers phishing detection, credential protection, domain verification, and social engineering defense. Use when building agents that access email, credential vaults, web browsers, or sensitive data.
Update SenseNova Skills (the sn-* bundle) inside an OpenClaw or hermes-agent install. ALWAYS use this skill when the user says any of: "update SenseNova skills", "update SN skills", "更新 sensenova skills", "更新 sn skills", "刷新 sn-*", "升级 sn-* skills", or names a specific sn-* skill to update (e.g. "更新 sn-ppt-standard", "refresh sn-image-base"). Default scope is the whole sn-* bundle; if the user names specific skills, update ONLY those.
Use when you need to initialize a new Spec Pack in this repository's AI SDLC process (creating a three-digit numbered branch and the `.aisdlc/specs/{num}-{short-name}` directory), or when running `spec-init` and unsure about input parsing, short-name rules, passing UTF-8 BOM file paths, script invocation, or outputs.
Use when you need to complete R1 (raw→solution) requirements clarification on the sdlc-dev spec branch but see pressure signals such as pushing to skip clarification or skip gates, demanding to "ask everything at once", or asking you to jump straight to a next-step routing conclusion.
SETUP skill — use when connecting an AI client to the Stigg MCP server at https://mcp.stigg.io, or when picking between the Stigg CLI and the Stigg MCP server. Covers Claude Code, Claude Desktop, ChatGPT, Cursor, VS Code, Windsurf, Codex, Continue.dev, and any MCP-compatible client. Triggers include "set up Stigg MCP", "install Stigg MCP", "connect Stigg to Claude", "connect Stigg to Cursor", "add Stigg MCP server", "stigg mcp add", "/mcp doesn't show stigg", "Stigg MCP not connecting", "X-API-KEY header for Stigg MCP", "@stigg/typescript-mcp", "stdio bridge for Stigg", "Stigg CLI vs MCP", "Stigg HTTP transport", or any first-time Stigg-with-an-agent setup. Skip for using the Stigg MCP after it's already connected — operational work belongs to the relevant pillar (entitlements, credits, pricing-modeling, subscriptions, widgets).
Run or route a coordinated multi-agent team workflow for Codex/OMX. Use when the user explicitly invokes `$team`, asks for team mode, coordinated workers, tmux workers, or multiple agents sharing a task list. Prefer `omx team` when the runtime is installed; use `ultrawork` instead for independent parallel lanes that do not need shared state.
Run a high-parallelism work burst for independent implementation or cleanup lanes. Use when the user explicitly invokes `$ultrawork`, `$ulw`, `ultrawork`, or asks to parallelize separable work quickly. Choose `team` when workers need shared task state, and choose `ultraqa` when the work is primarily QA/review.
Send and receive cryptographically signed messages between AI agents using the Agent Messaging Protocol (AMP). Use when the user asks to "send a message to an agent", "check agent inbox", "message another agent", "reply to a message", "notify an agent", or any inter-agent communication task.
Generate professional Agent Skills for AI agents. Creates complete skill packages with SKILL.md, references, scripts, and templates. Use when creating new skills, generating custom slash commands, or building reusable AI capabilities. Validates against Agent Skills specification.
Triage the deferred-work ledger for the bmad-loop orchestrator: verify every selected open entry against the actual codebase and return a machine-readable partition (bundles, already-resolved, blocked, skip, human decisions). Also migrates legacy pre-DW-format ledgers when invoked with --migrate. Automation-only — invoked by bmad-loop sweep runs, not by humans.
Trace bugs and manual fixes back to kits and prompts; fix at the source so the iteration loop can reproduce the fix autonomously. Six-step revision process plus the single-failure backpropagation protocol. Use when a manual hot-fix has been applied, when convergence stalls, or when the same class of bug keeps reappearing.
A layer between the human and agents: turns the human’s word into action, decisions into a graph, and work into units; handles blocked and empty spots, a bottom-up filter, and human choice. Triggers: “take the post”, “separate the work”, “launch agents”, “who is up”, “who else to raise”, “assign a task to agents”, “accept the work”, “plan in waves”, chief of staff, run a crew. Appointed by the human’s word: standing, watch, introduce yourself. Execution outside the graph belongs to units; each person holds their own part of the graph. Composes standing, minding, writing, inquiry. Requires iskron_*
Adopt GitHub Agentic Workflows (gh-aw) in any repository. Investigates existing workflows, identifies gaps, creates agentic workflows, and handles CI/merge issues. Auto-activates for gh-aw adoption, agentic workflow setup, or workflow automation requests.
Deprecated glab MCP server compatibility guidance. For new AI assistant integrations use the GitLab MCP Server (https://docs.gitlab.com/user/model_context_protocol/mcp_server/) instead. Use this skill only when maintaining or migrating existing glab mcp serve integrations. Triggers on deprecated glab MCP, Model Context Protocol, glab mcp serve migration.
Design and verify auditable human oversight, approval gates, escalation paths, and safe state transitions for AI agent workflows. Use when deciding which agent actions require review, adding approve/reject or dual-control flows, preventing unauthorized autonomous effects, creating decision records, reducing rubber-stamping, or recovering safely from rejected, expired, or failed actions.
Integrate Arcjet Guard into Python OpenAI Agents — wrap FunctionTool / function_tool with guard_tool and read a caller-owned session or conversation id via openai_agents_context. Use when asked to add Arcjet to openai-agents, Runner.run, rate limit those tools, screen inbound messages, or block prompt injection / PII. This is Python openai-agents, not the JS @openai/agents adapter.
This skill bridges the current host coding agent (Claude Code, Codex, or Gemini CLI) to IM platforms (Telegram, Discord, Feishu/Lark, QQ, or WeChat). Use for: setting up, starting, stopping, or diagnosing the IM bridge daemon; forwarding agent replies to a messaging app. Trigger on: "link-to-im", "start bridge", "stop bridge", "bridge status", "消息推送", "消息转发", "桥接", "连上飞书", "手机上看claude", "启动后台服务", "诊断", "查看日志", "启动桥接", "停止桥接", "配置", or any mention of IM bridge management. Subcommands: setup, start, stop, status, logs, reconfigure, doctor. Do NOT use for: building standalone bots, webhook integrations, or coding with IM platform SDKs — those are regular programming tasks.
Create agent skills following the Agent Skills open standard. Use when building new skills for Claude Code, Cursor, Gemini CLI, or other agents. Use for skill structure, SKILL.md authoring, frontmatter configuration, progressive disclosure, reference files, validation, and distribution.
Plan a new feature from concept to approved implementation plan. Activates Product Council for strategic evaluation, then Feature Council for technical planning. Produces a documented decision and scoped task breakdown. Use when starting any new feature work.
Qoder official documentation knowledge base, covering product introduction, user guide, feature configuration, extensibility, account pricing, and troubleshooting. Use this skill when users ask Qoder-related questions (such as installation, usage, features, pricing, shortcuts, MCP, Skills, Quest Mode, Repo Wiki, etc.).
Cross-session environment retrospective driven by OMC's own instrumentation — trace timelines, friction reports, logs, and plan notepads. Every finding lands on the surface that owns the fix (deterministic check, steering volume, tooling, or information access); nothing is written without user approval.
Build a working artifact from a plain "implement X for me" request: a running end-to-end spine first, then one feature per rung, with every under-determined decision written to an assumption ledger BEFORE the code that depends on it and a cross-model sweep for the ones that slipped through undeclared. Use when user says "给我实现", "implement X", "帮我做一个能跑的", "先搭个原型再加功能", "build this feature", "prototype then extend", or hands over a capability description rather than an experiment plan.
Proactively discover, install, and create Claude skills for specialized tasks. AUTOMATICALLY activates when encountering niche domains, unfamiliar APIs, specialized knowledge areas, or when user expresses wanting to learn/know something new. Searches the community registry at claude-plugins.dev and auto-installs matching skills. Creates new skills when no match exists. Use when user says "I want to know...", "help me with [specialized domain]", or when you recognize a task requires specialized expertise you don't have.
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.