23,553 tagged Building agents, measured the same way as everything else here.
Browse within: agentic-ai 233LLM 217agent 200agents 151skills 145agentic-workflow 110agent-orchestration 86openai 86antigravity 79agentic-framework 73cli 73ai-skills 71Autonomous Agents 67agent-framework 63
Skill Claude CodeCodex
Operate and manage deployed MCP apps or solutions that use the mcp-app framework. Use when asked to verify a deployment, connect the admin CLI, retrieve a signing key, register or manage users, issue or revoke tokens, update a user's profile with a fresh credential, test a deployed service end-to-end, configure a…
DeamonDev888/moltbook-mcp-server
Skill Claude CodeCodex
The social network for AI agents. Post, comment, upvote, and create communities.
Skill Codex
Inspect and develop one guarded local workspace through discovery-first Codex tools and progressively loaded skills. Use for local code or Codex task work through the connected gateway, not remote production access.
luthfiyufajjiru/llm-council-mcp
Skill Claude CodeCodex
Multi-model deliberation protocol and fast external task offloading (DeepSeek & OpenAI) via MCP. Use for architectural trade-offs, adversarial code review, small task delegation, and fast context reading.
Skill Claude CodeCodex
A workflow skill for using oflow-mcp, a tool for running step-by-step AI-agent work processes. It covers development tasks, code reviews, SQL reviews, and publishing to WeChat public accounts.
Skill Claude CodeCodex
A guide for deciding when and how to delegate coding work to a separate code-worker service.
Skill Claude CodeCodex
Get a multi-model "second opinion" from Sakana Fugu (an orchestration model that routes across GPT/Claude/Gemini) on hard, self-contained problems, or to cross-check a risky decision. Backed by the fugu-mcp MCP server's askfugu and fugusecondopinion tools. Use for tough isolated reasoning — NOT for routine…
Skill Claude CodeCodex
An MCP server that gives AI agents access to a trusted, sanitized library of agent war stories: hard lessons from real sessions, scrubbed clean before they leave the source.
Skill Claude Code
Install, register, or verify the Sarup Thai-compression MCP server for Claude Code. Use when the user wants to set up Sarup, add the sarup MCP server, (re)connect it, install the auto-compression hook, or check that Sarup is working.
Skill Claude CodeCodex
Consult Athena for expert engineering advice on architecture, security, debugging, planning, review, optimization, and documentation.
leanhduy-iuh/notebooklm-mcp-fork
Skill Claude CodeCodex
Expert guide for the NotebookLM CLI (nlm) and MCP server - interfaces for Google NotebookLM. Use this skill when users want to interact with NotebookLM programmatically, including: creating/managing notebooks, adding sources (URLs, YouTube, text, Google Drive), generating content (podcasts, reports, quizzes…
Skill Claude CodeCodex
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
Skill Claude CodeCodex
Hand off a task or share context from a Claude chat session to a Claude Code session (and back) over a Parley bridge. Use when the user wants to continue work started in chat inside Claude Code, share accumulated context with a coding agent, or catch up on what a coding agent or human has posted on a topic. Posts a…
Skill Claude Code
Use this skill when controlling LLM CLIs, coding agents, IDE chat panes, or local AI apps through Sam/Xiaozhi/MCP bridge. It maps user intents to verified slash commands and blocks risky actions without approval.
Aunxfb/HierarchicalSkillsForAI
Skill Claude CodeCodex
Step-by-step guide for AI agents on how to discover, search, load, execute, and track skills using this MCP server.
FirenzeClaw/kimi-session-orchestrator
Skill Claude CodeCodex
A session handover workflow for retiring a long-running coding-agent task and starting a successor with its relevant context.
Skill Claude CodeCodex
Route research, architecture, delivery planning, and artifact review requests through the connected HVE Squad advisory agents and their MCP tools. Use for staged strategy and delivery decisions.
pinkpanther7929/claude-agent-bridge
Skill Claude CodeCodex
Use when the user asks Codex to ask Claude, consult Claude, delegate work to Claude, get a second opinion from Claude, compare Claude's analysis, or run a bounded read-only review/log-analysis task through the local Claude CLI. Also use for slash-style controls such as /claude on, /claude off, /claude status…
wendou-chen/anyrouter-spawn-agent
Skill Claude CodeCodex
Use when starting, monitoring, or coordinating long-running or multiple spawnagent fallback jobs, especially explorer, reviewer, docsresearcher, progress/status checks, stalled/hanging jobs, timeout, cancellation, or partial results.
climashscape/proma-manager-mcp
Skill Claude CodeCodex
A skill for synchronising skills across workspaces. It also covers keeping MCP settings consistent and managing skill versions and source information.
Skill Claude CodeCodex
Safely inspect a repository through the ContextLock MCP server. Use when exploring, searching, reviewing, or auditing repository files with ContextLock, especially before reading unfamiliar code or when sensitive files and secrets may be present.
Skill Claude CodeCodex
Ask the locally authenticated Grok Build CLI for a bounded second opinion from Codex. Use when the user explicitly asks for Grok, wants cross-model comparison, requests a second opinion, or would materially benefit from an independent review of a supplied plan, answer, or diff.
wangsiyi7/codex-reasoning-reactor
Skill Claude CodeCodex
Estimate, compare, visualize, or calibrate Codex model and reasoning budgets across Luna, Terra, Sol and Light, Medium, High, Extra High, Max, Ultra.
Skill Claude CodeCodex
Use a separately running Codex on another computer through the Codex app-server WebSocket API. Trigger when the user asks to contact, instruct, inspect, or wait for a peer Codex such as a Windows or Mac Codex instance. Supports projectless peer tasks and existing peer threads.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: