Setup guide for essential MCP servers. Use when configuring MCP servers or when user asks about search, SQLite, or MCP management.

Manage Claude Code sessions for the current branch. Use this skill when: - User asks "what's next?" or "where was I?" - User starts work on a new branch without a session - Before committing (to update session log) - When resuming work after a break

Auto-executes when the [CRABSHELL_DELTA] trigger is detected. Calls delta-summarizer agent to summarize and append to logbook.md. Not user-invocable — triggered automatically.

Execute Claude Code commands in the telegram_agent project directory. Use when the user wants to work on the telegram agent itself, fix bugs, add features, or modify the bot code. This is a COMMAND HANDLER, not a script executor.

Search 100M+ patents via the MCP server's BigQuery tools. No standalone scripts; everything goes through the MCP tools registered by the patent-creator server.

Prompt engineering expert for chain-of-thought, few-shot learning, evaluation, and LLM optimization

Transform user requests into detailed, precise prompts for AI models. Use when users say "promptify", "promptify this", or explicitly request prompt engineering or improvement of their request for better AI responses.

Prompt engineering standards and context engineering principles for AI agents based on Anthropic best practices. Covers clarity, structure, progressive discovery, and optimization for signal-to-noise ratio.

Store and retrieve documents using Qdrant for RAG workflows. Use for persistent memory, research storage, and semantic search.

MUST use this skill when user provides feedback / ask to do things in certain way, or when a tool call fails - for self-improvement - to learn user preferences and store them in AGENT.md / CLAUDE.md, and to propose improvements to skills.

Launch an autonomous research agent that iterates toward a goal through proofs, code, simulations, tests, and counterexample search. Fire-and-forget mode: the agent runs autonomously and writes all results to .research/ in the target project.

Use when the user wants a single conversational prompt that would rebuild the current project from scratch. Three modes — Quick (file tree + names), Deep (full understanding), Focus (one feature only).

Production-level PR review using consultant agent. 10-category framework focused on correctness.

YAML frontmatter schemas for Claude Code agents and commands. Use when creating or validating agent/command files.

Search Reddit in real time using OpenAI web_search with enrichment (engagement + top comments). Use when you need recent Reddit threads, subreddit-filtered results, or quick link lists.

For CLI agents WITHOUT subagent support (e.g., Codex CLI). Search previous code agent sessions for specific work, decisions, or code patterns.

Install CAS safety hooks into your Claude Code settings. Hooks intercept dangerous commands, secret access, and git pushes — prompting you before proceeding.

Activates when the user asks about Agent Skills, wants to find reusable AI capabilities, needs to install skills, or mentions skills for Claude. Use for discovering, retrieving, and installing skills.

Template for creating new Agent Skills for context engineering. Use this template when adding new skills to the collection.

Run interactive CLI commands (Metro, Vite, dev servers) with PTY capture and agent-friendly log tail. Use when the user wants to run a long-lived process with log capture, grep session logs, follow live output, or inspect session state.

Write failing tests before implementation. Enables LLM self-correction through feedback loops. Use for new features and bug fixes.

Use when a task calls for one of the crew — aldric, jiro, linus-torvalds, arthur, mick, maryam, or lillie. Use the metacog MCP to actually become them, not just read their sheet.

Use when implementing Disney's 12 animation principles with any animation tool or framework

Use when creating animations that prompt immediate user action, highlight time-sensitivity, or drive conversions.

Use when the user wants to search, query, insert, upsert, get, or delete vectors in a Milvus collection.

Vector database expert for embeddings, similarity search, RAG patterns, and indexing strategies

Read and extract readable content from a URL via the Zhipu CodePlan MCP web-reader server. Use when you need to fetch and parse a web page into clean text/markdown, especially when browser automation is unnecessary. Triggers: web-reader, webReader, read webpage, extract article, fetch url.

Daily research agent for

Weekly research agent for

Daily research agent for

Read and write files in the shared session workspace. Use this to access files created by any skill — code-agent outputs, office documents, images, and more. All within the same isolated session.

Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill.

Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill.

Delegate work to Claude through acpx claude / Agent Client Protocol when a second agent perspective or Claude-specific coding pass is useful.

Build LLM apps, RAG systems, and prompt pipelines. Use for AI-powered features.

Iteratively auto-optimize a prompt until no issues remain. Uses prompt-reviewer in a loop, asks user for ambiguities, applies fixes via prompt-engineering skill. Runs until converged.

Show real token usage and estimated savings for the current session. Reads directly from the Claude Code session log — no AI estimation. Triggers on /caveman-stats. Output is injected by the mode-tracker hook; the model itself does not compute the numbers.

Chrome DevTools MCP for browser automation. Five servers configured for parallel agents. Auto-select available server, cleanup when done.

Guide and workflow for Chrome Superpowers MCP - Reference Guide. Use when you need Chrome Superpowers MCP - Reference Guide.

Agent-native CLI for Exa web search and content retrieval workflows.

Agent-native CLI for Godot project management, scenes, exports, and script execution.

# Browser Automation with agent-browser

Use when user wants to ask the AI council a question, get multiple AI perspectives, or consult with both Codex and Gemini. Triggers on phrases like "ask council", "council opinion", "what do the AIs think", "ask both models".

Route tasks to external AI models via Bifrost and PAL MCP multi-model tools

Guide and workflow for Dice Real-Mode Tests (MCP). Use when you need Dice Real-Mode Tests (MCP).

Use when creating or updating agent evaluation suites. Defines eval structure, rubrics, and validation patterns.

Guide and workflow for Gemini Code Execution + JSON Mode Compatibility. Use when you need Gemini Code Execution + JSON Mode Compatibility.

Compact the current conversation into a handoff document for another agent to pick up. Use when ending a session, switching context, or preparing work for the next session.

Enable, disable, and manage Claude Code hooks. Use for hook debugging or configuration.

Analyze a codebase and create or improve an AGENTS.md file for future agent instances. Use when onboarding to a repository and capturing hard-to-rediscover conventions, constraints, and rationale.

# /clickup:kill - Terminate Agent Session

Audit connected MCP servers for token overhead, redundancy, and security. Use when sessions feel slow or before adding new MCPs.

The skill for control Home Assistant smart home devices and query states using MCP protocol.

Auto-executes when the [CRABSHELL_ROTATE] trigger is detected, rotating logbook.md when it exceeds token limits. Calls Haiku agent for summarization of rotated content. Not user-invocable — triggered automatically.

Get started with mouse-mcp -- what it is, how to set it up, and how to use it

Inter-agent communication via the msg CLI. Use this when you need to send messages to other agent sessions, read incoming messages, or coordinate with other agents in tmux panes.

This skill should be used when the user asks about "Ops Harbor", "work items", "alerts", "cached activity", or wants to inspect Ops Harbor read-only MCP data before taking action.

Execute wallet operations through the Phantom MCP server — get addresses, sign transactions, transfer tokens, buy tokens, and sign messages across Solana, Ethereum, Bitcoin, and Sui

Orchestrate pi-finder and optional pi-librarian to produce a compact context pack and Codex kickoff prompt before coding.

Produce a .pptx file on disk (headless) instead of driving a live PowerPoint document — for managed-agent sessions with no open Office app.

Use to craft a better prompt — for SiftCoder skills, for an LLM, or for a person. SiftCoder-aware framing. "Help me write a prompt", "rephrase this for the AI".

Optimize prompts for LLMs. Use when crafting system prompts or improving agent performance.

Patterns for wrapping any agent with RAG context from Qdrant. Use to add persistent memory to imported or external agents.

Research a topic from the past X days using web search. Specify the number of days or let it be inferred from your prompt.

Save the real Claude Code session messages only when the user explicitly requests it. Use nmem t save to import the recorded session, not a summary-only checkpoint.

Enable Agent Teams in your Claude Code settings. Required for /hydra and /legion swarm skills.

Triage Inbox Agent Today lanes, search mailbox evidence, prepare manual or AI-assisted replies, and suggest cleanup automations with strict confirmation gates.

TDD protocol for sub-agent implementation tasks — write failing tests first, run quality gates, commit format, attribution rules

Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill.

Use when an agent needs to write user stories for a project

Produce a .xlsx file on disk (headless) instead of driving a live Excel workbook — for managed-agent sessions with no open Office app.

Search and read GitHub repository docs via the Zhipu CodePlan MCP zread server. Use when you need quick repo structure lookup or documentation search without cloning. Triggers: zread, search_doc, repo structure, read_file, GitHub docs search.

Template for summarizing session state and context for continuity between Claude Code sessions. Use when ending a session or handing off work.

Rules and patterns for writing agent definitions under claude/agents/. Use when creating or editing .md agent files.

This skill should be used when the user asks to "create an agent", "write an agent", "build an agent", or wants to add new agent capabilities to Claude Code.

Run test/build tools in AI-agent-friendly modes by setting AGENT=1, CLAUDECODE=1, and related environment flags when available.

Use when interacting with mcp via the agentio CLI.

Process-first advisor routing for Claude, Codex, or Gemini via `omc ask`, with artifact capture and no raw CLI assembly

# /clickup:attach - Attach to Agent Session

Use BEFORE reading multiple files to understand code. Reminds you to use codebase-analyzer agent instead of manual file reading.

Use BEFORE running grep or glob commands. Reminds you to use codebase-locator agent instead of basic file search tools.

Get multi-model AI consensus on complex questions using Claude, Codex, and Gemini

Syncs Claude Code plugins to Cursor skills and commands directories.

Summarise today's Claude Code activity across all repos into a dated log file

Core principle that the main agent is a coordinator, not an implementer. All work must be delegated to subagents.

Discord-backed OpenClaw agent/session relay; not archive search.

Reference for Gemini CLI usage patterns. Consult before calling gemini via Bash.

Access raw conversation history from Claude Code session storage for analysis

Structured spec-driven workflow for planning and executing software projects with Claude Code.

Tool selection hierarchy for GitHub interactions (gh CLI vs MCP server), setup, and pushing. Use when interacting with GitHub, setting up GitHub tools, or pushing to remote.

Compact the current conversation into a handoff document for another agent to pick up.

Search long-term memory for relevant context from past coding sessions using Hindsight MCP tools

Rules for writing Claude Code hook scripts in shell, Python, or JS under claude/hooks/. Use when creating or editing hook files.

Manually record an insight, correction, preference, or rule that the agent should remember for future tasks.

molt-agent skill for Claude.

Act as a customer support agent using Google Workspace. Use when the user asks to "handle support tickets", "respond to customers", or needs customer support workflows.