31,977 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 agents7,557Claude Code6,620Data and AI3,682Backend and APIs3,496Languages3,401Planning3,217Git and workflow3,165Code review2,733Testing2,364Memory and context2,364DevOps and cloud2,310Writing and docs2,303Research2,297Frontend2,252Security2,013Productivity1,710
All Claude Code31,977 Codex23,157 OpenCode22,109 GitHub Copilot3,551 Gemini CLI1,917 Cursor1,739 Windsurf129 Kiro79
| Project | Stars | Configures | Files | Always loaded |
|---|---|---|---|---|
| OasAIStudio/open-agent-sdk openagentsdk.dev Lightweight, general-purpose TypeScript agent runtime. Open-source alternative to Claude A | 46 | Claude Code, Codex, OpenCode | 2 | 1,372 tok |
| bug-ops/mcp-execution 🚀 Generate executable TypeScript tools from MCP servers with 98% token savings. Progressi | 46 | Claude Code, GitHub Copilot | 14 | 19,355 tok |
| ralforion/orionbelt-analytics ralforion.com Ontology-based MCP server that analyzes database schemas (PostgreSQL, Snowflake, ClickHous | 46 | Claude Code, Codex, OpenCode | 2 | 1,606 tok |
| bitovi/convey bitovi.github.io Point-n-click vibe coding in your browser! Draw features, improve designs, report bugs. | 46 | Claude Code, Codex, GitHub Copilot | 21 | 2,187 tok |
| Agent-Hellboy/mcp-server-fuzzer A generic mcp server fuzzer | 46 | Claude Code, Codex, OpenCode | 2 | 1,406 tok |
| aaronjmars/tweazy tweazy.wtf The best way to monetize AI applications & MCP, using x402, CDP Smart Wallets & Paymaster | 46 | Claude Code | 1 | 813 tok |
| madsnorgaard/drupal-agent-resources Reusable Claude Code resources for Drupal development - skills, agents, and commands | 46 | Claude Code | 13 | 556 tok |
| konamgil/mandu An open-source framework that prevents AI agents from breaking your codebase. | 46 | Claude Code, Codex, OpenCode | 3 | 1,442 tok |
| hex0xdeadbeef/claude-kit | 46 | Claude Code | 38 | 6,576 tok |
| LingyiChen-AI/OmniKube Multi-cluster Kubernetes management platform supporting RBAC access control, resource mana | 46 | Claude Code | 1 | 734 tok |
| lowtidebuild/public-equity-research Institutional-grade stock research for US and Korean equities — delivered in minutes. | 46 | Claude Code, Codex, OpenCode | 14 | 11,721 tok |
| Anantys-oss/koan koan.anantys.com Autonomous background agent that consumes unused Claude Max quota to explore, audit, and i | 46 | Claude Code, Codex, GitHub Copilot, OpenCode | 18 | 7,998 tok |
| goldbergyoni/golden-testing-rules-for-ai Instructions for LLMs on how to produce great tests with best practices inside | 46 | Claude Code | 1 | 561 tok |
| thebnbrkr/agora-code Persistent memory and codebase intelligence for AI coding agents, indexes symbols, tracks | 46 | Claude Code | 12 | 743 tok |
| zauberzeug/odrive-gui GUI to configure the ODrive Motor Controller | 46 | Claude Code, Codex, OpenCode | 2 | 778 tok |
| DelineaXPM/delinea-mcp MCP server for the Delinea Secret Server and Platform APIs | 46 | Claude Code, Codex, OpenCode | 2 | 2,032 tok |
| HazAT/pi-solo Pi extension for Solo (soloterm.com) — registers every Solo MCP tool as a first-class Pi t | 46 | Claude Code, Codex | 1 | — |
| timothywarner/copilot-dev techtrainertim.com GitHub Copilot for Developers (Live Training Course) | 46 | Claude Code, Codex, GitHub Copilot, OpenCode | 16 | 8,592 tok |
| topheman/effect-viz effect-viz.vercel.app Effect runtime visualizer 🌀 | 46 | Claude Code, Codex, OpenCode | 2 | 1,006 tok |
| hachej/boring-ui Build agent-powered apps without reinventing the shell. | 46 | Claude Code, Codex, OpenCode | 77 | 1,322 tok |
| seamus-brady/springdrift springdrift.ai A persistent runtime for long-lived LLM agents | 46 | Claude Code | 1 | 25,304 tok |
| sorafujitani/ccsession pick any past Claude Code session via fzf, then resume it from the right cwd | 46 | Claude Code, Codex | 2 | — |
| gl0bal01/malware-analysis-claude-skills gl0bal01.com Complete Claude skills toolkit for professional malware analysis. 5 specialized skills cov | 46 | Claude Code | 1 | 1,697 tok |
| grandcamel/JIRA-Assistant-Skills Claude Code Skills for JIRA automation - modular skills for issue management, workflows, s | 46 | Claude Code | 1 | 5,871 tok |
| duceum/appstore-ppp-pricing-agent-skill pypi.org PPP pricing for App Store apps — bulk-update IAP & subscription prices in 175+ countries b | 46 | Claude Code | 1 | 1,751 tok |
| haabe/mycelium AI made building cheap. It didn't make deciding cheap. Mycelium is a Claude Code harness t | 46 | Claude Code, Codex, OpenCode | 3 | 5,773 tok |
| hamr0/barebrowse Gives agents a real browser. URL in, pruned snapshot out. Replaces Playwright, Selenium, P | 46 | Claude Code | 1 | 534 tok |
| NERVsystems/llm9p LLM exposed as a 9P filesystem | 46 | Claude Code | 1 | 2,810 tok |
| bardiakhosravi/tenets npmjs.com Architecture guardrails for AI coding agents building backend services with DDD and Hexago | 46 | Claude Code | 1 | 48 tok |
| CarlosVallejoRuiz/slurp Token-budget-aware graph navigation for AI coding agents. Serve exactly the noodles your L | 46 | Claude Code, Codex | 3 | 12,466 tok |
| ed-donner/personal-space | 46 | Claude Code, Codex, OpenCode | 6 | 845 tok |
| shelby/shelby-skills shelby.xyz Agent skills for building on Shelby | 46 | Claude Code, Codex, OpenCode | 2 | 404 tok |
| andrzejchm/fdb Flutter Debug Bridge - CLI for AI agents to interact with running Flutter apps on device | 46 | Claude Code, Codex, OpenCode | 8 | 1,534 tok |
| Naridon-Inc/aura auravcs.com An audit trail for what AI agents changed in your codebase, and why — signed, per commit, | 46 | Claude Code, GitHub Copilot, Cursor, Gemini CLI, Windsurf | 5 | 4,687 tok |
| martinkup/symfony-profiler-optimization-advisor-bundle Symfony Profiler meta data collector — reads from Doctrine, Cache, Twig, Events, HTTP Clie | 46 | Claude Code | 5 | 1,441 tok |
| crystal-autobot/autobot crystal-autobot.github.io Ultra-efficient personal AI assistant powered by Crystal | 46 | Claude Code | 5 | 286 tok |
| iSerter/laravel-claude-agents Awesome Claude Code Subagents and Skills for Laravel. | 46 | Claude Code | 1 | 174 tok |
| Randroids-Dojo/skills A collection of Claude Code skills by Randroids Dojo | 46 | Claude Code, Codex, OpenCode | 2 | 917 tok |
| vercel/microfrontends vercel.com Build microfrontends on Vercel | 46 | Claude Code, Codex, OpenCode | 2 | 2,188 tok |
| TheAstrelo/Claude-Pipeline Tool-agnostic 13-phase AI development pipeline — turns a task description into reviewed, c | 46 | Claude Code, Codex, OpenCode | 40 | 7,476 tok |
| joosure/Maestro A platform evolution of OpenAI Symphony (openai/symphony) for running Codex, Claude Code, | 46 | Claude Code, Codex | 6 | — |
| asdhabdua/bilibili-video-notes-skill Bilibili Video Notes is a tool that automatically generates DOCX study notes with screensh | 46 | Claude Code, Codex, OpenCode | 2 | 1,817 tok |
| more-io/claude-apple-bridges Swift CLI bridges giving Claude Code native access to Apple Reminders, Calendar, Contacts, | 46 | Claude Code | 1 | 2,005 tok |
| tothalex/drift tothalex.github.io Review your working changes like a pull request — in the terminal. Block-scoped live diffs | 46 | Claude Code, Codex, OpenCode | 2 | 1,340 tok |
| JubaKitiashvili/everything-react-native-expo erne.dev Complete AI coding agent harness for React Native and Expo development | 46 | Claude Code | 58 | — |
| rusch95/asahi_neo Asahi Linx on a Macbook Neo Experimental | 46 | Claude Code | 1 | 613 tok |
| eai-org/agent-toolkit eai-org.github.io Minimalistic, project-agnostic skills and rules for AI coding agents | 46 | Claude Code, Codex, OpenCode | 2 | 253 tok |
| shuymn/gh-mcp A GitHub CLI extension that seamlessly runs the github-mcp-server using your existing gh a | 46 | Claude Code | 1 | 925 tok |
| neuromechanist/research-skills neuromechanist.github.io A skill marketplace for academic research: from project management to literature review, m | 46 | Claude Code, Codex, OpenCode | 2 | 3,332 tok |
| benchjack/benchjack AI agent benchmark hackability scanner — find evaluation vulnerabilities before they under | 46 | Claude Code | 2 | 614 tok |