awesome-ai-coding

Curated AI coding tools, plugins, and skills to accelerate software development.
https://github.com/youxufkhan/awesome-ai-coding

Last synced: about 8 hours ago
JSON representation

⚙️ AI Agent Frameworks & Multi-Agent Platforms
- Agency Agents - agents?style=flat-square&label=%E2%98%85)](https://github.com/msitarzewski/agency-agents) | msitarzewski | Specialized expert agent pack for creative agencies. | Social media bots, copy editors, UX reviewers, community monitors. |
- Agno - agi/agno?style=flat-square&label=%E2%98%85)](https://github.com/agno-agi/agno) | Agno AI | Open-source framework for building multi-modal agents. | Fast execution, built-in vector integrations, multi-modal support. |
- AI Berkshire - berkshire?style=flat-square&label=%E2%98%85)](https://github.com/xbtlin/ai-berkshire) | xbtlin | A value investing research framework built on Claude Code using multi-agent adversarial analysis. | Python / Claude Code |
- AI Hedge Fund - hedge-fund?style=flat-square&label=%E2%98%85)](https://github.com/virattt/ai-hedge-fund) | virattt | Simulated multi-agent hedge fund team. | Agent roles: market analyst, risk manager, portfolio optimizer, trader. |
- AI Scientist v2 - Scientist-v2?style=flat-square&label=%E2%98%85)](https://github.com/SakanaAI/AI-Scientist-v2) | SakanaAI | Fully automated scientific discovery agent. | Tree-search planning, paper writing, data generation, automated peer review. |
- AutoGen - square&label=%E2%98%85)](https://github.com/microsoft/autogen) | Microsoft | Multi-agent conversation framework. | Customizable agent personalities, human-in-the-loop, conversable structures. |
- AutoGPT - Gravitas/AutoGPT?style=flat-square&label=%E2%98%85)](https://github.com/Significant-Gravitas/AutoGPT) | Significant Gravitas | One of the earliest autonomous agent projects. | Multi-agent pipelines, web search, memory integration. |
- BabyAGI - square&label=%E2%98%85)](https://github.com/yoheinakajima/babyagi) | Yohei Nakajima | Lightweight task-driven agent loop. | Simple agent architecture utilizing vector databases and task management pools. |
- CrewAI - square&label=%E2%98%85)](https://github.com/crewAIInc/crewAI) | CrewAI Inc | Role-based agent teams orchestration. | Easy agent role-play definitions, task delegation, sequential or hierarchical execution. |
- DeepAgents - ai/deepagents?style=flat-square&label=%E2%98%85)](https://github.com/langchain-ai/deepagents) | LangChain AI | Batteries-included agent harness library. | Standard task interfaces, tool execution wrappers, built-in debug trackers. |
- Deer Flow - flow?style=flat-square&label=%E2%98%85)](https://github.com/bytedance/deer-flow) | ByteDance | Long-horizon SuperAgent harness for automated research. | Deep memory architecture, sandboxed execution, goal decomposition. |
- Dexter - square&label=%E2%98%85)](https://github.com/virattt/dexter) | virattt | Autonomous agent for deep financial research. | Market scraping, portfolio modeling, automated report drafting. |
- Generic Agent - square&label=%E2%98%85)](https://github.com/lsdefine/GenericAgent) | lsdefine | Self-evolving AI agent with dynamic skill trees. | Seeded by a 3.3K line Python script, builds its own skills, low token overhead. |
- Hermes Agent - agent?style=flat-square&label=%E2%98%85)](https://github.com/NousResearch/hermes-agent) | Nous Research | An agent that learns and evolves its internal loop processes. | Self-correction heuristics, modular agent graph, adaptive tool binding. |
- LangGraph - ai/langgraph?style=flat-square&label=%E2%98%85)](https://github.com/langchain-ai/langgraph) | LangChain AI | Stateful multi-agent graph orchestrator. | Cycles, human-in-the-loop, structured state, production-ready. |
- LobeHub - square&label=%E2%98%85)](https://github.com/lobehub/lobehub) | LobeHub | Chief Agent Operator for organizing and scheduling multi-agent teams. | Multi-agent collaboration, 7x24 agent operations, deep client integration. |
- MetaGPT - square&label=%E2%98%85)](https://github.com/FoundationAgents/MetaGPT) | FoundationAgents | Multi-agent framework mimicking software teams. | Simulates product managers, designers, and developers via SOPs. |
- MiroFish - square&label=%E2%98%85)](https://github.com/666ghj/MiroFish) | 666ghj | Swarm intelligence prediction engine. | Multi-node consensus, predictive optimization loops. |
- Multica - ai/multica?style=flat-square&label=%E2%98%85)](https://github.com/multica-ai/multica) | Multica AI | Managed open-source agent platform. | Skill compounding, real-time progress logging, teammate assign systems. |
- OpenMontage - square&label=%E2%98%85)](https://github.com/calesthio/OpenMontage) | calesthio | Open-source agentic video production system with 12 pipelines, 52 tools, and 500+ agent skills. | Python / Video Generation |
- Page Agent - agent?style=flat-square&label=%E2%98%85)](https://github.com/alibaba/page-agent) | Alibaba | JavaScript in-page GUI agent that controls web interfaces with natural language. | TypeScript / Web GUI Control |
- ruflo - square&label=%E2%98%85)](https://github.com/ruvnet/ruflo) | ruvnet | Leading agent meta-harness for Claude to deploy multi-agent swarms. | Swarm intelligence, adaptive memory, self-learning workflows, RAG integration. |
- Sim - square&label=%E2%98%85)](https://github.com/simstudioai/sim) | Sim Studio | Workforce agent management layer. | Integrates multi-agent tasks, live team logging, worker synchronization. |
- TradingAgents - square&label=%E2%98%85)](https://github.com/TauricResearch/TradingAgents) | TauricResearch | Multi-agent LLM financial trading framework. | Simulated trading environments, multi-agent portfolios. |
🤖 AI Coding CLI Agents & SWE Systems
- Aider - chat/aider?style=flat-square&label=%E2%98%85)](https://github.com/aider-chat/aider) | Aider-chat | Git-integrated terminal pair programmer. | Edits files in-place, writes automatic git commits, works with local/remote LLMs. |
- Claude Code - line agent powered by Claude models. | 1M+ token context window, deep reasoning, multi-file code workspace editing. |
- Claw Code Example - code?style=flat-square&label=%E2%98%85)](https://github.com/ultraworkers/claw-code) | Ultraworkers | Rust-based repository managed automatically by agentic code. | Demonstrates zero human intervention maintenance, autonomous documentation. |
- cc-switch - switch?style=flat-square&label=%E2%98%85)](https://github.com/farion1231/cc-switch) | farion1231 | Cross-platform desktop All-in-One assistant manager for Claude Code, Codex, OpenCode, OpenClaw, Gemini CLI & Hermes. | Desktop wrapper, multi-agent provider switching, WSL support. |
- Devika - square&label=%E2%98%85)](https://github.com/mufeedvh/devika) | Mufeed VH | Open-source clone of Cognition's Devin. | Plan-execute-monitor coding loops, agentic browser search. |
- Gemini CLI - gemini/gemini-cli?style=flat-square&label=%E2%98%85)](https://github.com/google-gemini/gemini-cli) | Google Gemini | An open-source, terminal-first AI agent powered by Gemini. | ReAct loops, MCP server compatibility, Google Search grounding, git workflows. |
- Goose - goose/goose?style=flat-square&label=%E2%98%85)](https://github.com/aaif-goose/goose) | Agentic AI Foundation (AAIF) | Extensible open-source coding agent. | Local or desktop run modes, built-in MCP client, custom developer plugins. |
- GPT-Engineer - engineer-org/gpt-engineer?style=flat-square&label=%E2%98%85)](https://github.com/gpt-engineer-org/gpt-engineer) | GPT-Engineer Org | Full codebase generator from prompts. | Generates complete frameworks, interactive query staging. |
- Mentat - square&label=%E2%98%85)](https://github.com/ggarrett8/mentat) | Mentat | Context-aware codebase editor. | Customized file indexing, fast command loop, interactive diff views. |
- Oh My Codex (OmX) - Heo/oh-my-codex?style=flat-square&label=%E2%98%85)](https://github.com/Yeachan-Heo/oh-my-codex) | Yeachan Heo | Orchestration hooks and head-up display tools for local coders. | Multi-agent configurations, diagnostic heads, scriptable workflow bindings. |
- oh-my-openagent - yeongyu/oh-my-openagent?style=flat-square&label=%E2%98%85)](https://github.com/code-yeongyu/oh-my-openagent) | code-yeongyu | A plugin and agent harness for OpenCode that orchestrates multiple models into specialized agents. | CLI coding agent harness, plugin integrations. |
- Oh My Pi - my-pi?style=flat-square&label=%E2%98%85)](https://github.com/can1357/oh-my-pi) | can1357 | A terminal-native AI coding agent with advanced tool harness. | Hash-anchored edits, LSP integration, browser automation, subagent delegation. |
- Open Agent (AFK-surf) - surf/open-agent?style=flat-square&label=%E2%98%85)](https://github.com/AFK-surf/open-agent) | AFK-surf | Open-source alternative to proprietary agents like the Claude Agent SDK or ChatGPT Agents, emphasizing multi-agent collaboration. | TypeScript-based, multi-agent collaboration, computer/browser/telephone operations. |
- OpenAgent - open-agent/openagent?style=flat-square&label=%E2%98%85)](https://github.com/the-open-agent/openagent) | the-open-agent | Next-generation personal AI assistant with agent loops, RAG, and computer/browser control for coding tasks. | LLM-powered computer/browser control, RAG, custom agent loops. |
- OpenAgentsControl - square&label=%E2%98%85)](https://github.com/darrenhinde/OpenAgentsControl) | Darren Hinde | Plan-first developer agent framework with code review, testing, and approval-based execution. | Multi-language support (TypeScript, Python, Go), review-driven approvals. |
- OpenAI Codex CLI - cookbook?style=flat-square&label=%E2%98%85)](https://github.com/openai/openai-cookbook) | OpenAI | Basic terminal execution wrappers for Codex/GPT-4. | Script writing, direct terminal-based shell translations. |
- Open Claude Code Decompile - claude-code?style=flat-square&label=%E2%98%85)](https://github.com/ruvnet/open-claude-code) | ruvnet | Reverse-engineered and rebuilt nightly Claude Code CLI helper. | Custom hooks, custom execution routing, diagnostic extensions. |
- OpenClaw - square&label=%E2%98%85)](https://github.com/openclaw/openclaw) | OpenClaw | Personal AI assistant running on any platform. | Agentic task orchestration, cross-platform local execution, visual tools. |
- OpenClaude - square&label=%E2%98%85)](https://github.com/Gitlawb/openclaude) | Gitlawb | Adaptable terminal assistant designed to run on any model provider. | Flexible tool bindings, de-coupled backend, command automation. |
- OpenCode - square&label=%E2%98%85)](https://github.com/anomalyco/opencode) | Anomaly | Model-agnostic terminal developer agent. | Go-based TUI, 75+ LLM providers supported, LSP diagnostics, version control snapshots. |
- OpenHands (formerly OpenDevin) - Hands-AI/OpenHands?style=flat-square&label=%E2%98%85)](https://github.com/All-Hands-AI/OpenHands) | All-Hands AI | Docker-sandboxed autonomous coding agent. | Runs shell commands, browses the web, reads/writes code inside isolated containers. |
- OpenSWE - ai/open-swe?style=flat-square&label=%E2%98%85)](https://github.com/langchain-ai/open-swe) | LangChain AI | Open-source asynchronous coding agent framework. | Stateful planning loops, async task execution, customizable LLM backend routes. |
- Pi Agent - works/pi?style=flat-square&label=%E2%98%85)](https://github.com/earendil-works/pi) | Earendil Works | A terminal-first AI agent toolkit and harness. | Unified LLM API, agent loop, interactive TUI, session persistence, modular tool architecture. |
- Plandex - ai/plandex?style=flat-square&label=%E2%98%85)](https://github.com/plandex-ai/plandex) | Plandex AI | Large-task codebase editor. | Sandboxed file staging, handles massive refactoring pipelines. |
- PR Agent - PR-Agent/pr-agent?style=flat-square&label=%E2%98%85)](https://github.com/The-PR-Agent/pr-agent) | Qodo / CodiumAI | Autonomous PR reviewer and description generator. | Git provider integrations (GitHub, GitLab, Bitbucket), auto-docs, code suggestions. |
- Rowboat - square&label=%E2%98%85)](https://github.com/rowboatlabs/rowboat) | Rowboat Labs | Local-first AI coworker that builds a persistent, structured knowledge graph from files and notes. | Markdown context storage with backlinks, situational awareness for agent workflows. |
- SWE-agent - nlp/SWE-agent?style=flat-square&label=%E2%98%85)](https://github.com/princeton-nlp/SWE-agent) | Princeton NLP | Agent-Computer Interface (ACI) coder framework. | Top SWE-bench performer, optimized file editing and search tools for LLMs. |
- Sweep - square&label=%E2%98%85)](https://github.com/sweepai/sweep) | Sweep AI | GitHub issue solver bot. | Automatic PR creation, lint fixing, and CI/CD validation. |
🧩 AI Coding Extensions & Plugins
- GitHub Copilot
- Continue - square&label=%E2%98%85)](https://github.com/continuedev/continue) | VS Code, JetBrains | Continue | Open-source, model-agnostic extension. Connects to Ollama (local) or API keys. |
- Cline - square&label=%E2%98%85)](https://github.com/cline/cline) | VS Code | Cline | Open-source agent interface. Reads/writes files and runs terminal commands via MCP. |
- Roo Code - Code?style=flat-square&label=%E2%98%85)](https://github.com/RooVetGit/Roo-Code) | VS Code | Roo Vet | Cline fork with specialized agent modes, prompt overrides, and MCP diagnostics. |
- Cody - with-docs. |
- Amazon Q Developer
- Supermaven - low latency autocomplete, large 300,000 token context window. |
- Codeium
- Double.bot - accuracy inline editing, codebase context parsing. |
- Tabnine
- Bito
🛡️ AI Gateways, Proxies & Agent Security
- 9router - square&label=%E2%98%85)](https://github.com/decolua/9router) | Decolua | Local proxy and router for AI tools. | Connects CLI agents to 40+ providers, automatic fallbacks, token compression. |
- Clawshield - public?style=flat-square&label=%E2%98%85)](https://github.com/SleuthCo/clawshield-public) | SleuthCo | YAML-driven security gateway for dev agents. | Enforces prompt policies, logs outputs, filters outbound credentials. |
- CodeGate - square&label=%E2%98%85)](https://github.com/stacklok/codegate) | Stacklok | Privacy-first local proxy for IDE extensions/CLIs. | Detects/masks PII and API keys, warns of vulnerable packages, local decryption. |
- Kiji Proxy - proxy?style=flat-square&label=%E2%98%85)](https://github.com/dataiku/kiji-proxy) | Dataiku | Lightweight PII masking proxy. | Regular expression and NLP-based token masking. |
- LiteLLM - square&label=%E2%98%85)](https://github.com/BerriAI/litellm) | BerriAI | Call 100+ LLM APIs using the OpenAI format. Proxy server for load balancing, cost tracking, and fallbacks. | Load balancer, budget manager, billing tracking, fallback routing. |
- OmniRoute - square&label=%E2%98%85)](https://github.com/diegosouzapw/OmniRoute) | diegosouzapw | AI gateway supporting smart routing, load balancing, retries, and fallback mechanisms for multi-provider LLMs. | OpenAI-compatible endpoint, retries, multi-provider fallbacks. |
- One API - api?style=flat-square&label=%E2%98%85)](https://github.com/songquanpeng/one-api) | songquanpeng | Multi-provider LLM API key management, distribution, and billing proxy system. | OpenAI API format wrapper, multi-user quotas, custom billing. |
- Portkey AI Gateway - AI/gateway?style=flat-square&label=%E2%98%85)](https://github.com/Portkey-AI/gateway) | Portkey-AI | High-performance, open-source AI gateway to route, load-balance, cache, and monitor LLM queries. | 50+ LLM integrations, fast routing, caching, observability. |
- RouteLLM - sys/RouteLLM?style=flat-square&label=%E2%98%85)](https://github.com/lm-sys/RouteLLM) | lm-sys | Open-source framework for cost-saving LLM routing, sending simpler queries to smaller models and complex queries to larger ones. | Cost optimization, router training and evaluation, BERT-based routing models. |
- WitnessAI - grade AI proxy firewall. | Granular policy configurations, real-time agent auditing, schema obscuring. |
📖 AI Learning & Educational Resources
- AI Agents for Beginners - agents-for-beginners?style=flat-square&label=%E2%98%85)](https://github.com/microsoft/ai-agents-for-beginners) | Microsoft | 12-lesson roadmap for learning to build AI agents. | Structured lectures, notebook labs, tool-calling and planning basics. |
- Claude Code Best Practice - code-best-practice?style=flat-square&label=%E2%98%85)](https://github.com/shanraisshan/claude-code-best-practice) | shanraisshan | Guidelines and tips for transition from vibe coding to agentic engineering with Claude Code. | Structured best practices, prompt optimization, workflow patterns. |
- Learn Claude Code - lab/learn-claude-code?style=flat-square&label=%E2%98%85)](https://github.com/shareAI-lab/learn-claude-code) | shareAI-lab | Educational nano-agent tutorial showing how to build a Claude Code-like harness from scratch. | Code walkthroughs, lightweight bash-to-LLM bindings, agentic tutorials. |
- Modded NanoGPT - nanogpt?style=flat-square&label=%E2%98%85)](https://github.com/KellerJordan/modded-nanogpt) | Keller Jordan | Extremely fast, modified NanoGPT training repo. | Achieves 124M parameter training in 90 seconds, showcases optimization. |
- Neural Networks: Zero to Hero - zero-to-hero?style=flat-square&label=%E2%98%85)](https://github.com/karpathy/nn-zero-to-hero) | Andrej Karpathy | Detailed guide to building neural networks from scratch. | From basic backpropagation to GPT, PyTorch labs, video lecture companion. |
- Prompt Engineering Guide - ai/Prompt-Engineering-Guide?style=flat-square&label=%E2%98%85)](https://github.com/dair-ai/Prompt-Engineering-Guide) | DAIR.AI | Curated guides, papers, notebooks, and learning paths for prompt/context engineering and AI agents. | Structured papers index, learning materials, system prompt models. |
- Unsloth - square&label=%E2%98%85)](https://github.com/unslothai/unsloth) | Unsloth AI | Web UI and training environment for fine-tuning open weights models. | 2x-5x faster model training, lightweight local setup (Gemma, Qwen, DeepSeek). |
💻 AI-Native IDEs & Cloud Workspaces
- Cherry Studio - studio?style=flat-square&label=%E2%98%85)](https://github.com/CherryHQ/cherry-studio) | CherryHQ | AI productivity studio with smart chat, autonomous agents, and 300+ assistants. | Unified access to frontier LLMs, multiple assistant panels, offline capabilities. |
- ClawX - ai/ClawX?style=flat-square&label=%E2%98%85)](https://github.com/ValueCell-ai/ClawX) | ValueCell-ai | Graphical desktop interface for OpenClaw AI agents. | Visual agent configuration, no-code execution controls, local execution logs. |
- Cursor - file agent, autocomplete, inline codebase chat, custom index. |
- GitHub Copilot Workspace - centric cloud workspace. | Interactive issue plans, automated task lists, instant sandbox edits. |
- Melty - labs/melty?style=flat-square&label=%E2%98%85)](https://github.com/melty-labs/melty) | Melty Labs | Open-source IDE that learns from developers. | Tracks git actions, code pattern indexing, smart autocompletion. |
- PearAI - source VS Code fork. | Custom assistant panel, inline refactoring, privacy controls. |
- Project IDX - based developer workspace. | Multiplatform templates, integrated Gemini chat, device emulators. |
- Zed AI - performance, collaborative rust-based IDE. | Multi-buffer editing, instant startup, inline LLM chats (OpenAI/Anthropic). |
🔌 Custom Agent Skills & Plugins
- Academic Research Skills - research-skills?style=flat-square&label=%E2%98%85)](https://github.com/Imbad0202/academic-research-skills) | Imbad0202 | Workflows for research, compilation, drafting, and reviewing papers. | Claude Code / General Agents |
- Agent Skills - skills?style=flat-square&label=%E2%98%85)](https://github.com/addyosmani/agent-skills) | Addy Osmani | Production-grade software engineering skills. | Generic / Claude Code, Gemini CLI |
- Andrej Karpathy Skills - ai/andrej-karpathy-skills?style=flat-square&label=%E2%98%85)](https://github.com/multica-ai/andrej-karpathy-skills) | Multica AI | Claude Code behaviors derived from Karpathy's LLM coding pitfall observations. | Claude Code / CLAUDE.md |
- Anthropic Cybersecurity Skills - Cybersecurity-Skills?style=flat-square&label=%E2%98%85)](https://github.com/mukul975/Anthropic-Cybersecurity-Skills) | mukul975 | 817 structured cybersecurity skills for AI agents mapped to MITRE, NIST frameworks. | Claude Code / General Agents |
- Awesome Claude Skills - claude-skills?style=flat-square&label=%E2%98%85)](https://github.com/ComposioHQ/awesome-claude-skills) | ComposioHQ | Curated list of awesome Claude Skills, resources, and tools. | Claude Code / General Agents |
- Caveman - square&label=%E2%98%85)](https://github.com/JuliusBrussee/caveman) | Julius Brussee | Token-saving skill that rewrites instructions into caveman talk. | Claude Code / General Agents |
- ClawHub - square&label=%E2%98%85)](https://github.com/openclaw/clawhub) | OpenClaw | Skill and plugin registry for OpenClaw agents. | OpenClaw |
- DESIGN.md - labs-code/design.md?style=flat-square&label=%E2%98%85)](https://github.com/google-labs-code/design.md) | google-labs-code | Format specification for describing a visual identity to coding agents. DESIGN.md gives agents a persistent, structured understanding of a design system. | Generic / Coding Agents |
- Graphify - square&label=%E2%98%85)](https://github.com/safishamsi/graphify) | safishamsi | Turn code, schemas, scripts, docs, and media into queryable knowledge graphs. | Claude Code / Codex / Gemini CLI / Cursor |
- gstack - square&label=%E2%98%85)](https://github.com/garrytan/gstack) | Garry Tan | Garry Tan's exact Claude Code setup: 23 opinionated tools that serve as CEO, Designer, Eng Manager, Doc Engineer, and QA. | Claude Code / CLAUDE.md |
- Guizang PPT Skill - ppt-skill?style=flat-square&label=%E2%98%85)](https://github.com/op7418/guizang-ppt-skill) | op7418 | AI-agent Skill for generating polished HTML slide decks with Swiss layouts. | Claude Code / Codex / General Agents |
- jscpd - square&label=%E2%98%85)](https://github.com/kucherenko/jscpd) | Andrey Kucherenko | Copy/paste detector for programming source code, AI-ready with token-efficient reporter and skill. | Generic / Coding Agents |
- Last 30 Days Skill - skill?style=flat-square&label=%E2%98%85)](https://github.com/mvanhorn/last30days-skill) | M. Van Horn | Scrapes and synthesizes web/forum summaries. | Claude Code / Generic Agents |
- Obsidian Skills - skills?style=flat-square&label=%E2%98%85)](https://github.com/kepano/obsidian-skills) | Kepano | Teaches agents to read Markdown, Bases, and JSON Canvas in Obsidian. | Obsidian / Generic Agents |
- React Doctor - doctor?style=flat-square&label=%E2%98%85)](https://github.com/millionco/react-doctor) | Million.co | Catch and fix bad React code generated by AI agents. | Generic / Code Review Agents |
- Superpowers - square&label=%E2%98%85)](https://github.com/obra/superpowers) | Obra | An agentic skills framework & software development methodology that works. | Generic / Claude Code, Gemini CLI |
- Taste-Skill - skill?style=flat-square&label=%E2%98%85)](https://github.com/Leonxlnx/taste-skill) | Leonxlnx | Taste-Skill - gives your AI good taste, stopping boring and generic layout generation. | Claude Code / Codex |
🛠️ LLM DevTools & Orchestration Infrastructure
- CodeGraph - square&label=%E2%98%85)](https://github.com/colbymchenry/codegraph) | Colby McHenry | Pre-indexed local code knowledge graph. | Syncs automatically on git changes, exposes code indexes locally to avoid API calls. |
- ClipCannon - square&label=%E2%98%85)](https://github.com/ChrisRoyse/clipcannon) | Chris Royse | AI-driven video processing tool that maps transcripts and media segments into semantic vector clusters. | Teleological constellations mapping, video content vector indexing, semantic search. |
- Dify - square&label=%E2%98%85)](https://github.com/langgenius/dify) | LangGenius | Visual LLM app building and operations platform. | Workflow orchestrator, prompt engineering UI, vector db integrations, monitoring dashboards. |
- Flowise - square&label=%E2%98%85)](https://github.com/FlowiseAI/Flowise) | FlowiseAI | Drag & drop UI for building LangChain/LlamaIndex apps. | Visual nodes, instant API endpoints, chat widget deployments. |
- GitNexus - square&label=%E2%98%85)](https://github.com/abhigyanpatwari/GitNexus) | Abhigyan Patwari | Client-side, zero-server repository intelligence engine. | Runs in-browser, maps code dependencies, outputs AI-friendly knowledge maps. |
- Heretic - e-w/heretic?style=flat-square&label=%E2%98%85)](https://github.com/p-e-w/heretic) | p-e-w | Automatic model censorship removal tool. | Local model tuning layer, automated system alignment bypass. |
- LangChain - ai/langchain?style=flat-square&label=%E2%98%85)](https://github.com/langchain-ai/langchain) | LangChain AI | De facto standard ecosystem for LLM orchestration. | Chains, prompts, data connectors, custom tool integrations, LangSmith observability. |
- Lightpanda Browser - io/browser?style=flat-square&label=%E2%98%85)](https://github.com/lightpanda-io/browser) | Lightpanda | Headless browser built in Zig for AI agents. | High-speed, lightweight DOM execution, optimal for sandboxed agent scraping. |
- LlamaFactory - square&label=%E2%98%85)](https://github.com/hiyouga/LlamaFactory) | hiyouga | Unified efficient fine-tuning framework for 100+ LLMs and VLMs. | Web UI dashboard, multi-GPU training support, RLHF/QLoRA methods. |
- LlamaIndex - llama/llama_index?style=flat-square&label=%E2%98%85)](https://github.com/run-llama/llama_index) | LlamaIndex | Data framework for LLM/RAG applications. | Data ingestors, indexing tools, structured queries, custom query engines. |
- Mem0 - square&label=%E2%98%85)](https://github.com/mem0ai/mem0) | Mem0 | Personalized memory layer for AI agents. | Remembers user preferences, updates context over time, optimizes personalization. |
- MinerU - square&label=%E2%98%85)](https://github.com/opendatalab/MinerU) | OpenDataLab | Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for agentic workflows. | High-precision PDF layout extraction, structural element recognition. |
- n8n - io/n8n?style=flat-square&label=%E2%98%85)](https://github.com/n8n-io/n8n) | n8n-io | Fair-code workflow automation platform with native AI capabilities. | Visual workflow builder, 400+ integrations, native agent/subagent orchestration nodes. |
- OpenDataLoader PDF - project/opendataloader-pdf?style=flat-square&label=%E2%98%85)](https://github.com/opendataloader-project/opendataloader-pdf) | OpenDataLoader | Open-source PDF parser optimized for LLM training. | High accuracy structure mapping, table extraction, cleans layout noise. |
- OpenViking - square&label=%E2%98%85)](https://github.com/volcengine/OpenViking) | Volcengine | Context database tailored for AI agents. | Persistent agent memory storage, runtime resource management, skill indexes. |
- PaddleOCR - square&label=%E2%98%85)](https://github.com/PaddlePaddle/PaddleOCR) | PaddlePaddle | Document-to-text OCR parser. | Supports 100+ languages, extracts layout and tables from scanned documents. |
⚡ LLM Inference & Local Serving Engines
- GPT4All - ai/gpt4all?style=flat-square&label=%E2%98%85)](https://github.com/nomic-ai/gpt4all) | Nomic AI | Desktop app and SDK for local LLM execution. | Runs locally on consumer GPUs/CPUs, built-in search/document RAG capabilities. |
- LiteRT-LM - ai-edge/LiteRT-LM?style=flat-square&label=%E2%98%85)](https://github.com/google-ai-edge/LiteRT-LM) | Google AI Edge | Edge device model inference engine. | Highly optimized for mobile devices and embedded hardware. |
- llama.cpp - org/llama.cpp?style=flat-square&label=%E2%98%85)](https://github.com/ggml-org/llama.cpp) | GGML Org | High-performance inference engine in C/C++. | CPU/GPU execution, low memory overhead, broad hardware support. |
- Llamafile - Ocho/llamafile?style=flat-square&label=%E2%98%85)](https://github.com/Mozilla-Ocho/llamafile) | Mozilla Ocho | Single-file executable for local LLMs. | Runs across 6 OS platforms without installations, fast CPU execution. |
- llmfit - square&label=%E2%98%85)](https://github.com/AlexsJones/llmfit) | AlexsJones | Find out which local models and providers run on your hardware. | Model hardware scanner, benchmark testing, single-command usage. |
- LocalAI - square&label=%E2%98%85)](https://github.com/mudler/LocalAI) | mudler | Drop-in OpenAI replacement for local models. | Supports audio generation, images, embedding, and text models. |
- Ollama - square&label=%E2%98%85)](https://github.com/ollama/ollama) | Ollama | Package and run LLMs locally on macOS, Linux, and Windows. | Simple CLI, one-line installs (Llama 3, Phi, Qwen), OpenAI compatible API. |
- SGLang - project/sglang?style=flat-square&label=%E2%98%85)](https://github.com/sgl-project/sglang) | SGLang Project | Structured generation language & engine. | High throughput serving, prompt caching, JSON schema constraints. |
- Text-generation-webui - generation-webui?style=flat-square&label=%E2%98%85)](https://github.com/oobabooga/text-generation-webui) | oobabooga | Gradio web UI for running local models. | Supports transformers, llama.cpp, ExLlamaV2, custom parameter sliders. |
- vLLM - project/vllm?style=flat-square&label=%E2%98%85)](https://github.com/vllm-project/vllm) | vLLM Project | High-throughput, memory-efficient LLM serving engine. | PagedAttention, continuous batching, distributed inference, serving API. |
📡 Model Context Protocol (MCP) Servers
- Agent Toolkit for AWS - toolkit-for-aws?style=flat-square&label=%E2%98%85)](https://github.com/aws/agent-toolkit-for-aws) | AWS | Official AWS-supported MCP servers, skills, and plugins to help AI agents build on AWS. | AWS services access (S3, EC2, Bedrock, etc.) |
- Claude Context MCP - context?style=flat-square&label=%E2%98%85)](https://github.com/zilliztech/claude-context) | Zilliz Tech | Vector search-based code search MCP server. | Code search over huge codebases, fast local similarity query tools. |
- Context7 - square&label=%E2%98%85)](https://github.com/upstash/context7) | Upstash | Serverless state and tool cache. | Caches tool executions, manages state across agent handoffs. |
- GitHub MCP Server - mcp-server?style=flat-square&label=%E2%98%85)](https://github.com/github/github-mcp-server) | GitHub | Official GitHub API bridge. | Lists issues, manages PRs, commits changes, searches repositories. |
- MCP Servers Repo - square&label=%E2%98%85)](https://github.com/modelcontextprotocol/servers) | Anthropic | Standard repository of reference MCP servers. | Postgres, SQLite, Slack, GitHub, Puppeteer, Brave Search. |
- Playwright MCP - mcp?style=flat-square&label=%E2%98%85)](https://github.com/microsoft/playwright-mcp) | Microsoft | Browser automation server. | Exposes headless browser actions (navigate, click, type, screenshot). |
🗄️ Vector Databases & Vector Stores
- Chroma - core/chroma?style=flat-square&label=%E2%98%85)](https://github.com/chroma-core/chroma) | Chroma Core | Lightweight, developer-friendly open-source vector store. | Fast python setup, ideal for notebooks and rapid prototyping. |
- Deep Lake - square&label=%E2%98%85)](https://github.com/activeloopai/deeplake) | activeloopai | AI Data Runtime for Agents, serverless postgres, and multimodal datalake. | Multimodal storage, RAG integration, vector query and training runtime. |
- FAISS - square&label=%E2%98%85)](https://github.com/facebookresearch/faiss) | Meta Research | Library for efficient similarity search of dense vectors. | GPU acceleration, fast index algorithms, local run implementation. |
- Milvus - io/milvus?style=flat-square&label=%E2%98%85)](https://github.com/milvus-io/milvus) | Milvus | High-performance, distributed vector database. | Built for billion-scale datasets, GPU indexing, hybrid search support. |
- Qdrant - square&label=%E2%98%85)](https://github.com/qdrant/qdrant) | Qdrant | Rust-based vector similarity search engine. | Fast payload filtering, REST/gRPC interfaces, production scale. |
🎨 Vibe Coding & Generative UI Tools
- AFFiNE - square&label=%E2%98%85)](https://github.com/toeverything/AFFiNE) | toeverything | Workspace tool unifying notes, tasks, and graphics. | AI-native drawing, structured page layout generation. |
- AI Website Cloner Template - website-cloner-template?style=flat-square&label=%E2%98%85)](https://github.com/JCodesMore/ai-website-cloner-template) | JCodesMore | Clone any website with one command using AI coding agents. | TypeScript / Website cloning |
- Chatbot UI - ui?style=flat-square&label=%E2%98%85)](https://github.com/mckaywrigley/chatbot-ui) | McKay Wrigley | A clean, customizable front-end for LLM APIs. | Custom prompt templates, system instructions, attachment parsing. |
- InsForge - square&label=%E2%98%85)](https://github.com/InsForge/InsForge) | InsForge | Backend-as-a-service (BaaS) platform optimized for agentic coding. | PostgreSQL database, auth, storage, edge functions, structured schemas for AI agent usage. |
- NanoChat - square&label=%E2%98%85)](https://github.com/karpathy/nanochat) | Andrej Karpathy | Minimalist ChatGPT clone. | Simple code, easily extensible, fast API connections. |
- Onyx - dot-app/onyx?style=flat-square&label=%E2%98%85)](https://github.com/onyx-dot-app/onyx) | Onyx App | Open source chat platform designed for multiple LLMs. | Advanced chat options, workspace customization. |
- OpenUI - square&label=%E2%98%85)](https://github.com/wandb/openui) | Weights & Biases | Generates responsive HTML/CSS layouts visually from description. | Direct Tailwind/React component export, interactive design editor. |
- Thunderbolt - square&label=%E2%98%85)](https://github.com/thunderbird/thunderbolt) | Thunderbird | Personal client for running customizable AI models. | Local data control, no vendor lock-in UI. |

Programming Languages

Python 33 TypeScript 32 C++ 7 Go 5 JavaScript 4 Jupyter Notebook 4 Rust 4 HTML 2 Shell 2 MDX 2

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Awesome

awesome-ai-coding

⚙️ AI Agent Frameworks & Multi-Agent Platforms

🤖 AI Coding CLI Agents & SWE Systems

🧩 AI Coding Extensions & Plugins

🛡️ AI Gateways, Proxies & Agent Security

📖 AI Learning & Educational Resources

💻 AI-Native IDEs & Cloud Workspaces

🔌 Custom Agent Skills & Plugins

🛠️ LLM DevTools & Orchestration Infrastructure

⚡ LLM Inference & Local Serving Engines

📡 Model Context Protocol (MCP) Servers

🗄️ Vector Databases & Vector Stores

🎨 Vibe Coding & Generative UI Tools