AI coding & agents · Weekly GitHub chart, Wed, Nov 18
AI coding & agents repositories pooled from every language chart and re-ranked by stars gained. As of Wed, Nov 18, 6 AM GMT+9.
- 1
mattpocock/skills
Composable engineering skills for AI coding agents
A collection of small, composable skills for coding agents such as Claude Code, Codex, GitHub Copilot, and Gemini CLI. It provides repeatable workflows for aligning requirements, TDD, debugging, and reviewing codebase architecture.
+6,658in 7 days
- 2
GoogleChrome/modern-web-guidance-src
Modern web platform skills for coding agents
This is the source repository for Modern Web Guidance, a set of skills that embed web platform expertise, best practices, and browser compatibility data into coding agents. It steers agents away from legacy patterns toward native modern web platform solutions.
+6,542in 7 days
- 3
anthropics/skills
Public repository for Claude Agent Skills
This repository provides Anthropic's implementation and examples of Agent Skills, which are folders of instructions, scripts, and resources that Claude loads dynamically. It covers creative, technical, enterprise, and document-handling workflows.
+5,666in 7 days
- 4
obra/superpowers
Agentic skills framework and software development methodology
Superpowers is a software development methodology for coding agents built on composable skills that trigger automatically. Instead of jumping straight into code, it guides agents through brainstorming, design approval, implementation planning, and subagent-driven execution.
+5,126in 7 days
- 5
milind-soni/OpenMausBot
Open-source desktop chat app for local AI agent teams
OpenMausBot is an open-source, local-first desktop app that lets you run a roster of AI bots powered by your existing Claude, Codex, and Grok CLIs inside a messaging interface. Each bot has its own model, a cloud or local virtual machine for computer use, and connected apps, with inline approval cards before executing sensitive actions.
+4,737in 7 days
- 6
vercel-labs/agent-desktop
Desktop automation CLI for AI agents
agent-desktop is a CLI and C library that lets AI agents inspect and operate desktop apps through the OS accessibility tree without screenshots or pixel matching. It returns UI elements as JSON with element refs so agents can click, type, and verify results.
+4,451in 7 days
- 7
Leonxlnx/taste-skill
Web design skill for coding agents
Taste Skill is an Agent Skill that guides coding agents to build company websites with a unified design system, real photographs, and browser verification instead of generic AI landing pages. It installs via the skills CLI or as a Claude Code or Codex plugin and runs on Node 22+ with Chrome, Edge, or Chromium.
+4,442in 7 days
- 8
google/elephant-goldfish-model
Open-source plugin for designing and reviewing software with agents
EGM is an open-source plugin for AI assistants where an Elephant role drafts specifications with project context and Goldfish reviewers examine them without prior conversation history to expose blind spots.
+4,249in 7 days
- 9
addyosmani/agent-skills
Production-grade engineering skills for AI coding agents
This repository packages engineering workflows, quality gates, and best practices so AI coding agents follow them consistently across every phase of software development. It covers the full lifecycle from defining specs and planning to building, testing, reviewing, and shipping.
+3,960in 7 days
- 10
vercel-labs/agent-browser
Browser automation CLI for AI agents
A native Rust CLI that lets AI agents drive a browser one shell command at a time. It reads pages as compact accessibility-tree text with element refs such as @e1 and acts on them while keeping the browser open between commands.
+3,760in 7 days
- 11
Alishahryar1/free-claude-code
Multi-provider proxy for AI coding agents
An open-source proxy that routes 12 AI coding harnesses, including Claude Code, Codex, and OpenCode, to models across 61 providers. It supports terminal, desktop app, IDE, phone, and native browser sessions.
+3,313in 7 days
- 12
alibaba/open-code-review
Hybrid deterministic and LLM code review CLI
An AI-powered code review CLI tool that reads Git diffs or full files to generate line-level review comments. It combines deterministic file selection and rule matching with an LLM agent for dynamic context retrieval.
+3,056in 7 days
- 13
blader/humanizer
Agent skill that removes AI writing patterns
Humanizer is an agent skill based on Wikipedia's 'Signs of AI writing' guide that rewrites AI-generated text to sound like a person wrote it. It preserves the original meaning and claims while removing common AI writing tells.
+2,860in 7 days
- 14
ChromeDevTools/chrome-devtools-mcp
Chrome DevTools MCP server for coding agents
An MCP server and CLI tool that lets coding agents such as Antigravity, Claude, Cursor, and Copilot control and inspect a live Chrome browser. It exposes Chrome DevTools capabilities for browser automation, debugging, and performance analysis.
+2,593in 7 days
- 15
tradermonty/claude-trading-skills
Claude Skills trading workflow toolkit for individual investors
A trading workflow toolkit for Claude Web and Claude Code that structures market review, risk management, trade planning, and journaling. Rather than automating orders or providing buy/sell signals, it focuses on supporting a trader's decision process.
+2,407in 7 days
- 16
Donchitos/Claude-Code-Game-Studios
Game development studio template for Claude Code
This project template configures Claude Code as a game development studio with 49 AI agents, 75 workflow skills, and 18 hook scripts. It guides you from a few design questions and a one-page brief to a playable browser prototype that can grow into a full project.
+2,381in 7 days
- 17
backnotprop/plannotator
Local review and annotation workspace for AI coding agents
Plannotator is a local browser-based review workspace where you can annotate plans, documents, HTML artifacts, and code diffs from AI coding agents and send feedback directly back to the agent. It connects through plugins, hooks, and MCP so agents can keep working while you review.
+2,325in 7 days
- 18
JuliusBrussee/caveman
Token-saving skill and proxy for coding agents
Caveman is a skill and proxy for AI coding agents that cuts token usage by shortening agent replies and compressing the data they read. It keeps code, commands, and file paths verbatim while trimming unnecessary prose.
+2,258in 7 days
- 19
homeassistant-ai/ha-mcp
MCP server for Home Assistant automation
An MCP server that lets AI assistants like Claude and ChatGPT configure and automate Home Assistant. It runs in-process as a Home Assistant integration or standalone via an app (add-on), Docker, or PyPI.
+2,221in 7 days
- 20
ranxianglei/billion-context
Context-compression plugin for long coding sessions
A proxy and plugin that sits between coding agents and model APIs to rewrite Anthropic and OpenAI streams with incremental context compression. Instead of hard truncation, the model calls built-in tools to compress, decompress, and search conversation history.
+2,023in 7 days
- 21
elementalsouls/Claude-BugHunter
Claude Code skill bundle for bug hunting and red-team work
A self-contained skill bundle for Claude Code designed for bug bounty hunting and external red-team engagements. It combines disclosed vulnerability report patterns with enterprise identity and infrastructure attack matrices.
+1,747in 7 days
- 22
ag-ui-protocol/ag-ui
Open event-based protocol connecting AI agents to frontend apps
An open, lightweight, event-based protocol that standardizes how AI agents connect to user-facing applications. It replaces custom wiring with a bi-directional event stream so any client can render agent outputs regardless of the underlying agent framework.
+1,569in 7 days
- 23
kajisho5/ffmpeg-skill
Local FFmpeg video editing skill for coding agents
An Agent Skill and MCP server that lets coding agents like Claude Code, Cursor, and Codex edit video with local FFmpeg and verify the output. It ships 42 structured tools backed by a machine-readable contract and runs locally without cloud services or API keys.
+1,528in 7 days
- 24
QwenLM/qwen-code
Open-source AI coding agent for terminal, editor, and web
Qwen Code is an open-source AI coding agent for terminals, desktop apps, web UI, IDEs, and chat integrations. It includes Auto-Memory, Skills, SubAgents, and MCP support alongside TypeScript, Python, and Java SDKs.
+1,497in 7 days
- 25
coreyhaines31/marketingskills
Marketing skills collection for AI coding agents
A collection of markdown-based skills that give AI coding agents specialized workflows for marketing tasks. It works with Claude Code, OpenAI Codex, Cursor, Windsurf, and other agents supporting the Agent Skills spec.
+1,418in 7 days