45,012 repositories keep their own agent configuration in the open: a CLAUDE.md, a rules file, instructions for Copilot, a skill the project runs on itself. These are not mods to install. They are worth reading for what they reveal — what a maintainer thinks an agent gets wrong.
Building agents8,549Claude Code6,896Git and workflow3,630Data and AI3,579Backend and APIs3,559Languages3,491Planning3,321Code review2,909Testing2,691DevOps and cloud2,492Writing and docs2,490Memory and context2,374Research2,321Frontend2,311Security2,183Project management1,977
All Claude Code32,540 Codex24,898 OpenCode23,819 GitHub Copilot3,723 Gemini CLI2,039 Cursor1,772 Windsurf202 Kiro116
| Project | Stars | Configures | Files | Always loaded |
|---|---|---|---|---|
| ttttonyhe/autogrind ✊ Let your agent grind on your projects 24x7 fully autonomously | 7 | Claude Code | 1 | 5,886 tok |
| H-Ali13381/tier-1-2-3-skill-system Tier 1/2/3 system for designing agent skills. | 7 | Codex, OpenCode | 1 | 195 tok |
| hellogxp/skill-runtime-intelligence Evidence-graded runtime panorama and diagnosis for Agent Skills | 7 | Codex, OpenCode | 1 | 270 tok |
| KevinRabun/judges MCP server with specialized judges to evaluate AI-generated code for security, cost, scala | 7 | GitHub Copilot | 5 | 6,029 tok |
| Marsmanleo/MarsNMe marsnme.com Agent-agnostic AI memory MCP — never forget a conversation again | 7 | Codex, OpenCode | 1 | 643 tok |
| armelhbobdad/oh-my-skills armelhbobdad.github.io Curated collection of agentskills.io-compliant, version-pinned agent skills — generated an | 7 | Claude Code | 1 | 1,154 tok |
| dmytri/shipshape skills.sh Shipshape is a context-isolated spec-driven development workflow for coding agents. Specs | 7 | Codex, OpenCode | 1 | 1,623 tok |
| deancourse/user-dot-agents Centralized version-controlled Claude Code user-level settings (skills / hooks / settings. | 7 | Claude Code | 1 | 522 tok |
| liyading1818/obsidian-llm-wiki Karpathy LLM Wiki for Obsidian: LLM-maintained knowledge base + Cursor agent skills (wiki, | 7 | Codex, OpenCode | 1 | 2,794 tok |
| snapsynapse/skill-provenance skillprovenance.dev Version identity and integrity verification for Agent Skills. Tracks versions, detects sta | 7 | Claude Code, Codex, Gemini CLI, OpenCode | 3 | 2,516 tok |
| altertable-ai/skills altertable.ai Agent Skills for Altertable | 7 | Codex, OpenCode | 1 | 1,813 tok |
| otto-haus/otto otto.haus Your agents forget every correction. otto makes them stick. | 7 | Codex, OpenCode | 1 | 2,449 tok |
| JasonCodeMaker/Agentic-Research-Control-Panel Local control plane for governed AutoResearch in real repos: human-approved scope, live ex | 7 | Claude Code, Codex, OpenCode | 2 | 1,894 tok |
| AgentiaPT/vela-slides agentiapt.github.io Vela Slides - AI Powered presentation App and Agent Skill | 7 | Claude Code | 4 | 5,642 tok |
| AytuncYildizli/yapitalism Drive terminal coding agents by voice, with receipts that never claim more than they prove | 7 | Claude Code, Codex | 1 | — |
| dasein108/yt-mem-ai pypi.org A local YouTube memory for your AI assistant — transcribe, remember, summarize. CLI + MCP | 7 | Claude Code, Codex, OpenCode | 2 | 4,919 tok |
| devrimcavusoglu/skern skern.dev Mass Skills Manager for Agents on AI Driven Development | 7 | Claude Code, Codex, OpenCode | 2 | 4,389 tok |
| iltempo/two-cents Unsolicited opinions for Claude Code — pick one and every answer gets an extra take nobody | 7 | Claude Code | 1 | 253 tok |
| alo-exp/silver-bullet sb.alolabs.dev Agentic Process Orchestrator (APO) — maximize AI-driven Dev Process Reliability at 10x low | 7 | Claude Code, Codex, OpenCode | 4 | 5,126 tok |
| lesliebiubiubiu/agent-context-kb-skill Lightweight, agent-facing routed knowledge base skill that gives coding agents durable pro | 7 | Claude Code, Codex, OpenCode | 2 | 802 tok |
| memi-design/design-sandbox memoire.cv memi-ready Next.js, Tailwind, shadcn, MCP, and Agent Skills sandbox for interface understa | 7 | Claude Code, Codex, OpenCode | 3 | 311 tok |
| janmaaarc/basecamp A production-ready setup for Claude Code and Codex. Shared coding rules, per-project memor | 7 | Claude Code, Codex, OpenCode | 2 | 7,100 tok |
| zhebil/claude-code-plugins Claude Code plugin marketplace with developer productivity hooks | 7 | Claude Code | 1 | 1,996 tok |
| DanielC000/loom danielc000.github.io Orchestrate a fleet of real Claude Code agents on your Claude subscription, not per-token | 7 | Claude Code | 1 | 28,615 tok |
| NEXPACE-Limited/msu-skills nexpace-limited.github.io Agent Skills for MapleStory Universe — a SKILL.md catalog for MSU builders. Works with Cla | 7 | Claude Code, Codex, Gemini CLI, OpenCode | 3 | 7,957 tok |
| normalled/apijack Generate full-featured CLIs with workflow automation from any OpenAPI spec | 7 | Claude Code, Codex, Gemini CLI, OpenCode | 3 | 7,648 tok |
| celanthe/clarion kiranoliver.com Your agents have things to say. Now they have a voice to say them with. | 7 | Claude Code, Codex | 22 | 1,385 tok |
| mzored/superpowers-cc-to-codex Claude orchestrates, Codex executes. Combines Claude's judgment with GPT's speed for codin | 7 | Claude Code, Codex, GitHub Copilot, Gemini CLI, OpenCode | 4 | 4,546 tok |
| vavasilva/ralph-planner Claude Code plugin that creates structured execution plans in Ralph Wiggum format | 7 | Claude Code | 1 | 577 tok |
| skullninja/coco-workflow Autonomous spec-driven development plugin for Claude Code. Describe a feature, get merged, | 7 | Claude Code | 2 | 9,612 tok |
| OrodruinLabs/nazgul Multi-agent autonomous development loop for Claude Code — 17 agents, review pipeline, cont | 7 | Claude Code | 2 | 13,150 tok |
| IS908/claude-lark-plugin claude code lark channel plugin | 7 | Claude Code | 1 | 4,085 tok |
| SpaiR/task-pipeline spair.github.io Interrogate the plan, capture what survived, implement it in any session — a chat-first ta | 7 | Claude Code | 3 | 2,343 tok |
| linxule/memex-plugin pypi.org Collaborative memory CLI and plugins for Claude Code, Codex, and Kimi Code: local vaults, | 7 | Claude Code | 1 | 3,580 tok |
| Dan-Cleary/convalytics convalytics.dev Simple analytics for Convex projects | 7 | Claude Code, Codex, OpenCode | 7 | 206 tok |
| OpenAnalystInc/10x-Canva-Skills 10x.in Claude Code skills plugin for Canva API integration and local file editing (PDF, PPTX, DOC | 7 | Claude Code | 1 | — |
| Phoenixrr2113/codebase-graph v0-landing-page-build-kappa-virid.vercel.app Code knowledge graph and MCP search server with tree-sitter parsing, vector + reranker ret | 7 | Claude Code | 1 | 3,376 tok |
| n0zer0d4y/chronos-protocol glama.ai A robust MCP server that eliminates temporal blindness in AI coding agents through intelli | 7 | Codex, OpenCode | 1 | 353 tok |
| yumeiriowl/repo-graphrag-mcp An MCP server that uses LightRAG and Tree-sitter to build a repository knowledge graph fro | 7 | Codex, OpenCode | 1 | 1,315 tok |
| eugeneyvt/logseq-mcp-server npmjs.com Model Context Protocol server for Logseq knowledge base integration with AI assistants | 7 | Claude Code, Codex, Cursor, OpenCode | 3 | 5,695 tok |
| marcomq/mq-bridge-app marcomq.github.io Bridge between multiple MQ message queues and streams like nats or kafka | 7 | Codex, OpenCode | 1 | 1,851 tok |
| ilia3546/swift-argument-parser-mcp A Swift library that turns any ArgumentParser CLI into an MCP server, letting AI agents c | 7 | Claude Code | 1 | 1,268 tok |
| steve-magne/hookstack hookstack.app Claude Code hooks catalogue — production-ready hooks for Claude Code, OpenAI Codex & GitHu | 7 | Claude Code, Codex, OpenCode | 28 | 9,006 tok |
| AndreasKarz/Vibe-Coding-Dev-Setup linkedin.com This repository contains a complete developer setup for .NET backend vibe coding with GitH | 7 | Claude Code, Codex, GitHub Copilot | 28 | 2,239 tok |
| daveangulo/twining-mcp npmjs.com Agent coordination MCP server — shared blackboard, decision tracking, and context assembly | 7 | Claude Code | 2 | 1,073 tok |
| sernnee/capacitor-mobile-claw Run a full AI agent on mobile with local LLM calls, on-device memory, code execution, and | 7 | Claude Code | 1 | 567 tok |
| lx-wnk/Agent-Context Layered, agent-agnostic context architecture for AI coding agents: a tiny always-on baseli | 7 | Claude Code | 1 | 1,475 tok |
| taniwhaai/arai arai.taniwha.ai Your AI assistant reads CLAUDE.md and ignores it anyway. Ārai makes instruction files enfo | 7 | Claude Code, Codex, OpenCode | 11 | 5,060 tok |
| mrosnerr/open-zk-kb Shared, persistent memory for AI assistants, built on the Zettelkasten method | 7 | Codex, OpenCode | 1 | 3,847 tok |
| ksaitor/teamcp teamcp.ai Self-hostable MCP gateway with per-member AI tool access control. | 7 | Claude Code | 1 | 1,768 tok |