| openclaw | ★ 390K | AI Agents | TypeScript | ai-assistant · multi-channel · self-hosted · local-first · plugins · skills · gateway · messaging-bots · privacy · sandboxing · cli · tui · nodejs · typescript · model-providers · personal-assistant | OpenClaw is a self-hosted, open-source AI assistant gateway that runs on your own devices and meets you in 20+ chat channels (Discord, Slack, WhatsApp, iMessage) plus native desktop and mobile apps. Models and agent harnesses are swappable plugins, with extensible tools, skills and a plugin SDK. | View Project → |
| superpowers | ★ 290K | Skills | Shell | agent-skills · skills-library · tdd · subagents · coding-agents · sdlc · methodology · brainstorming · code-review · git-worktrees · debugging · claude-code · cursor · codex · opencode · plugins | A cross-agent skills library and development methodology for coding agents, packaged as installable plugins. Skills auto-trigger brainstorming, plan writing, TDD, subagent-driven execution and code review across Claude Code, Codex, Cursor, Gemini CLI and others. | View Project → |
| skills | ★ 271K | Skills | Shell | agent-skills · claude-code · codex · spec-driven-development · tdd · code-review · debugging · context-engineering · adr · domain-modeling · issue-triage · prompt-engineering · workflow · plugin · shell · productivity | A set of small, composable agent skills for real software engineering with Claude Code, Codex, and other agents: grilling/alignment, spec-to-tickets, TDD, debugging, code review, triage, and architecture improvement. Installed via Claude Code plugin or npx skills, editable in-repo. | View Project → |
| ECC | ★ 269K | Skills | JavaScript | agent skills · coding agents · agent harness · memory · hooks · security · TDD · code review · research workflows · multi-platform | A cross-platform collection of agent skills, specialized agents, hooks, rules, and memory workflows for coding assistants. It adds structured development, review, security, and learning practices to Claude Code, Codex, Cursor, OpenCode, and other harnesses. | View Project → |
| hermes-agent | ★ 248K | AI Agents | Python | ai-agent · self-improving · tool-calling · agent-skills · memory · persistent-memory · multi-platform · messaging-gateway · subagents · cron-scheduler · agent-client-protocol · terminal-ui · self-hosted · model-agnostic · trajectory-generation · session-search | Hermes Agent is Nous Research's self-improving personal AI agent with tool calling, a built-in learning loop that writes and refines its own skills, persistent cross-session memory, and a messaging gateway for Telegram, Discord, Slack and more. Runs on CLI, VPS or serverless backends and is model-agnostic via Nous Port | View Project → |
| deepseek-harness | ★ 236K | AI Agents | TypeScript | agent-harness · plugin-architecture · cordis · deepseek · web-ui · desktop-app · typescript · nodejs · extensible · developer-preview · multi-platform · agent-lifecycle · api-gateway · persistence · monorepo · dsh-plugin | DeepSeek Harness (dsh) is DeepSeek AI's open-source agent harness built on an everything-is-a-plugin architecture via Cordis. It ships a local Web UI and desktop app, plus extensive docs on agent lifecycle, capability seams and plugin extension points. MIT, TypeScript, developer preview with breaking changes. | View Project → |
| opencode | ★ 211K | AI Coding | TypeScript | coding agent · AI coding · terminal · desktop app · code editing · codebase exploration · multi-agent · developer tool | Open-source AI coding agent available in the terminal and as a desktop app. Includes build and read-only planning agents, plus a general subagent for complex searches and tasks. | View Project → |
| n8n | ★ 206K | AI Agents | TypeScript | workflow-automation · no-code · low-code · ai-agents · mcp · mcp-client · mcp-server · self-hosted · integrations · visual-builder · human-in-the-loop · llm-orchestration · fair-code · templates · ipaas · typescript | n8n is a fair-code, self-hostable workflow automation platform with native AI: a visual canvas plus code for building multi-step AI agents, tool use, human approvals, and RAG, connected to 1500+ integrations and MCP servers/clients. | View Project → |
| tensorflow | ★ 200K | AI Frameworks | C++ | machine-learning · deep-learning · neural-networks · distributed-training · GPU · model-training · model-deployment | An end-to-end machine-learning framework with Python and C++ APIs for building, training, and deploying models across CPUs, GPUs, and other devices. | View Project → |
| claw-code | ★ 195K | AI Coding | Rust | rust · cli · agent-harness · mcp · acp · agent-skills · session-management · llm-cli · openai-compatible · docker · self-maintained · reference-implementation | Rust implementation of the `claw` CLI agent harness, published as an agent-managed museum exhibit rather than a production product. Offers prompt/session CLI, MCP plugin lifecycle, ACP status, Docker and Windows workflows; build-from-source only, with LazyCodex/Gajae-Code as the real harnesses. | View Project → |
| AutoGPT | ★ 188K | AI Agents | Python | autonomous-agents · agent-platform · workflow-automation · visual-builder · agent-marketplace · no-code · multi-agent · scheduling · triggers · self-hosted · docker · integrations · llm · python · agent-runtime · low-code | AutoGPT is an open-source platform for building, deploying, and running AI agents that execute full workflows. It offers a visual builder, plain-English AutoPilot agent creation, a community marketplace, and 45+ integrations, with both hosted and self-hosted options. | View Project → |
| markitdown | ★ 186K | AI Tools | Python | markdown · document-conversion · pdf · ocr · llm-preprocessing · rag · python · docx · pptx · xlsx · epub · youtube-transcription · audio-transcription · plugin-system · microsoft-office · text-extraction | Microsoft's lightweight Python utility that converts PDF, Office, image, audio, HTML, EPUB and YouTube content into LLM-friendly Markdown. Supports optional plugins, LLM-based image description/OCR, and Azure Document Intelligence or Content Understanding for higher-quality extraction. | View Project → |
| firecrawl | ★ 183K | AI Tools | TypeScript | web-scraping · web-crawling · markdown · llm · data-extraction · mcp · web-search · structured-data · browser-automation · api · crawler · html-to-markdown · ai-agents · pdf-parsing · cli · typescript | Firecrawl is an open-source web data API that searches, scrapes, crawls, and interacts with sites, returning clean markdown or structured JSON for LLMs. Ships SDKs, a CLI, an MCP server, and agent skills, and can also run an AI agent that gathers data from a prompt. | View Project → |
| ollama | ★ 181K | AI Frameworks | Go | llm · local-llm · inference · model-runner · llama.cpp · gguf · quantization · cli · rest-api · self-hosted · go · model-management · openai-compatible · embeddings · offline-ai · developer-tools | Ollama is a Go-based local LLM runtime that downloads, runs, and serves open models (Gemma, Qwen, DeepSeek, Llama) via a CLI and REST API on port 11434. It powers local inference for coding agents, chat UIs, and RAG apps, and supports custom Modelfiles and imports. | View Project → |
| skills | ★ 177K | Skills | Python | agent-skills · claude · skill-spec · prompt-engineering · skill-template · document-skills · claude-code-plugin · anthropic · skill-examples · marketplace | Anthropic's reference repository for Agent Skills: folder-based SKILL.md instruction packs that Claude loads dynamically. Includes example skills, the Agent Skills spec, a template, and installation via Claude Code plugin marketplace. | View Project → |
| transformers | ★ 166K | AI Frameworks | Python | transformers · model-definition · pretrained-models · llm · multimodal · nlp · computer-vision · audio · speech-recognition · inference · training · fine-tuning · pytorch · model-hub · vlm · pipeline-api | Hugging Face Transformers is the model-definition framework for state-of-the-art text, vision, audio, video and multimodal models. It offers a unified Pipeline/Trainer API over 1M+ Hub checkpoints and feeds model definitions to vLLM, llama.cpp and training stacks. | View Project → |
| stable-diffusion-webui | ★ 165K | AI Tools | Python | stable-diffusion · image-generation · txt2img · img2img · inpainting · outpainting · upscaling · gradio · web-ui · diffusion · textual-inversion · lora · face-restoration · prompt-engineering · pytorch · batch-processing | Gradio-based web UI for running Stable Diffusion locally: txt2img, img2img, inpainting, outpainting, upscaling and embedding/LoRA training, with a large extension ecosystem and REST API. Self-hosted on NVIDIA, AMD, Intel or Apple hardware. | View Project → |
| dify | ★ 157K | AI Frameworks | TypeScript | llm · agents · agentic-workflow · rag · low-code · no-code · workflow-builder · mcp · llmops · observability · prompt-ide · self-hosted · multi-model · baas · tool-calling · orchestration | Dify is an open-source LLM app platform combining a visual agentic workflow builder, RAG pipeline, prompt IDE, agent runtime, and LLMOps. It supports hundreds of models, MCP servers, and tools, and self-hosts via Docker with APIs for integration. | View Project → |
| langflow | ★ 155K | AI Frameworks | Python | visual-builder · low-code · multi-agent · workflows · orchestration · llm · mcp-server · rag · python · react-flow · api-deployment · playground · vector-databases · observability · self-hosted · no-code | Langflow is an open-source visual platform for building and deploying LLM agents and workflows. It combines a drag-and-drop React Flow canvas with Python component customization, then ships each flow as a REST API, an MCP server, or exportable JSON. | View Project → |
| agency-agents | ★ 154K | Skills | Shell | ai-agents · agent-personas · prompt-library · prompt-engineering · claude-code · cursor · codex · developer-tools · multi-agent · markdown-agents · installer-script · workflow-automation | A curated roster of ~100+ markdown agent personas (engineering, design, academic, marketing) with identity, process and deliverables, plus shell scripts to install them into Claude Code, Cursor, Codex, Gemini CLI and other AI coding tools. | View Project → |
| open-webui | ★ 153K | AI Tools | Python | self-hosted · llm-ui · chat-ui · ollama · openai-compatible · rag · mcp · agents · multi-model · image-generation · voice · rbac · pwa · vector-database · plugins · offline | Open WebUI is a self-hosted, extensible AI chat platform that front-ends Ollama and any OpenAI-compatible API. It bundles local RAG over 9 vector DBs, MCP/OpenAPI tool plugins, custom model presets and agents, voice/video, image generation, and enterprise RBAC/SSO. | View Project → |
| claude-code | ★ 148K | AI Coding | TypeScript | claude · agentic-coding · terminal · cli · developer-tools · code-assistant · git-workflows · ide-integration · plugins · anthropic · natural-language · codebase-understanding | Anthropic's official agentic coding tool that runs in your terminal or IDE, reads your codebase, and executes tasks, git workflows, and code explanations from natural-language prompts. Includes plugin support for custom commands and agents. | View Project → |
| langchain | ★ 147K | AI Frameworks | Python | agents · llm · rag · multi-agent · tool-calling · orchestration · prompt-templates · chat-models · embeddings · vector-stores · structured-output · streaming · integrations · model-interoperability · python · typescript | LangChain is the Python/JS framework for building LLM apps and agents: a standard interface for chat models, embeddings, vector stores, tools, and retrievers, plus chains, structured output, and streaming. Paired with LangGraph for controllable agent workflows. | View Project → |
| ponytail | ★ 144K | Skills | JavaScript | agent-skills · claude-code · claude-code-plugin · cursor-rules · prompt-engineering · yagni · developer-tools · llm · code-minimalism · agent-hooks · plugin-marketplace · code-review · cost-optimization · copilot · codex · gemini-cli | Ponytail is a cross-agent skill/plugin that makes coding agents apply a YAGNI ladder (skip, reuse, stdlib, native, one-liner) before writing code. Its benchmarks claim ~54% less code and ~20% lower cost on Claude Code sessions, while keeping validation, security, and accessibility intact. Ships Claude Code, Codex, Copi | View Project → |
| system-prompts-and-models-of-ai-tools | ★ 144K | Skills | — | system-prompts · prompt-engineering · ai-agents · ai-coding · agent-tools · reference · llm · cursor · claude-code · github-copilot · devin · windsurf · prompt-injection · ai-ides · v0 · replit | A large curated archive of leaked system prompts, tool/function schemas, and model configs from popular AI tools and coding agents (Cursor, Claude Code, Devin, Windsurf, v0, Perplexity and more). Useful as a reference for prompt design, agent tool schemas, and prompt-security research. | View Project → |
| awesome-llm-apps | ★ 139K | AI Agents | Python | agents · multi-agent · rag · llm · python · agent-skills · templates · tutorials · streamlit · voice-agents · browser-automation · always-on-agents · open-source · example-apps | Curated monorepo of 100+ runnable AI agent, agent-skill, and RAG app templates in Python, each with code and step-by-step tutorials. Good starting point if you want working examples to clone, run, and adapt with Claude, Gemini, GPT, or open models. | View Project → |
| spec-kit | ★ 138K | AI Coding | Python | spec-driven-development · ai-coding-agent · agent-skills · slash-commands · templates · prd · workflow · cli · github-copilot · extensions · presets · bundles · planning · developer-tooling | GitHub Spec Kit is a Python CLI plus agent skill packs that give AI coding agents a structured spec-driven workflow: define a spec, plan, tasks, then implement and converge, with optional bug-fixing and idea-assessment extensions. | View Project → |
| cc-switch | ★ 138K | AI Tools | Rust | claude-code · codex · gemini-cli · provider-management · api-key-management · mcp · skills-management · desktop-app · tauri · cross-platform · opencode · openclaw · grok · multi-provider · config-switcher · wsl | Tauri-based cross-platform desktop app that centrally manages, switches, and edits provider configs, API keys, MCP servers, and skills for AI coding agents (Claude Code, Codex, Gemini CLI, OpenCode, OpenClaw, Grok). Useful for developers juggling multiple LLM providers or relays. | View Project → |
| ComfyUI | ★ 134K | AI Tools | Python | diffusion-models · image-generation · video-generation · node-graph · workflow · text-to-image · image-editing · stable-diffusion · comfyui · controlnet · lora · inpainting · audio-generation · 3d-generation · api · local-inference | ComfyUI is a node-graph GUI, API, and backend for running diffusion models locally. It supports image, video, audio, and 3D generation workflows with hundreds of open models, quantized execution, and reusable JSON workflows. | View Project → |
| gstack | ★ 134K | Skills | TypeScript | claude-code · skills · slash-commands · ai-coding-agents · code-review · qa-automation · headless-browser · security-audit · release-automation · prompt-engineering · workflow-automation · developer-productivity · owasp · stride · typescript · bun | Garry Tan's Claude Code skill pack: 23 slash-command roles (CEO, eng manager, designer, reviewer, QA, security, release) plus CLI power tools like a headless browser, PDF maker, and diagram renderer. Installs into ~/.claude/skills; also supports Codex, Cursor, and other agents. | View Project → |
| ui-ux-pro-max-skill | ★ 130K | Skills | Python | ai-skills · ui-design · ux · design-system · design-tokens · color-palettes · typography · landing-page · claude-code · cursor · windsurf · copilot · tailwindcss · react · frontend · accessibility | An installable AI skill that turns a project brief into a full UI/UX design system: layout pattern, style, colors, fonts, effects, anti-patterns and a pre-delivery checklist, backed by 192 industry reasoning rules and multi-domain search. | View Project → |
| llama.cpp | ★ 129K | AI Frameworks | C++ | llm-inference · gguf · quantization · local-llm · cpu-inference · gpu-inference · cpp · openai-compatible-api · cuda · metal · vulkan · multimodal · edge-inference · no-dependencies · inference-engine · model-conversion | llama.cpp is a dependency-free C/C++ LLM and VLM inference engine built on ggml. It runs quantized GGUF models on CPU, CUDA, Metal, Vulkan and many other backends, and ships CLI, server, web UI and OpenAI-compatible API tools. | View Project → |
| llama.cpp | ★ 129K | AI Frameworks | C++ | C++ · llama.cpp · Inference | High-performance C/C++ inference for local LLMs across CPUs, GPUs, and devices. | View Project → |
| codex | ★ 126K | AI Coding | Rust | coding-agent · cli · terminal · rust · openai · sandbox · mcp · agent-skills · slash-commands · code-execution · non-interactive-exec · full-auto · ide-integration · chatgpt-auth · apache-2.0 · agent-policy | Codex CLI is OpenAI's lightweight open-source coding agent that runs locally in your terminal, written in Rust. It reads and edits code, runs shell commands in a sandbox, supports MCP servers, skills, and scriptable `codex exec`, and can sign in with a ChatGPT plan or API key. | View Project → |
| MoneyPrinterTurbo | ★ 125K | AI Tools | Python | AI video generation · short videos · video automation · text-to-speech · subtitles · script generation · content creation · WebUI · API · CLI · batch processing | Generates short videos from a topic or keyword by automating script writing, media selection, text-to-speech, subtitles, and editing. Offers a WebUI, API, and CLI, with support for multiple LLM providers and custom assets. | View Project → |
| graphify | ★ 121K | Skills | Python | Python · ai-agents · antigravity · ast · claude-code | Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store. | View Project → |
| generative-ai-for-beginners | ★ 120K | AI Tools | Jupyter Notebook | generative-ai · llm · prompt-engineering · rag · vector-database · ai-agents · fine-tuning · embeddings · chat-apps · image-generation · function-calling · responsible-ai · semantic-search · education · azure-openai · jupyter-notebook | Microsoft's 21-lesson course for developers learning to build generative AI apps. Covers LLM basics, prompt engineering, chat and search apps, RAG with vector databases, function calling, image generation, agents, fine-tuning, and responsible AI with Python and TypeScript samples. | View Project → |
| browser-use | ★ 116K | AI Agents | Python | browser-automation · ai-agents · computer-use · playwright · web-scraping · llm · python · browser-agent · task-automation · autonomous-agents · cli · mcp · structured-output · captcha-solving · multi-model · headless-browser | Browser Use is a Python library that lets LLM agents drive a real browser via Playwright to complete multi-step web tasks such as booking, form filling, and scraping. Supports OpenAI, Anthropic, Gemini, Groq, and local Ollama models, with custom tools, CLI/skill integration, and optional cloud browsers. | View Project → |
| whisper | ★ 110K | AI Tools | Python | speech-recognition · asr · transcription · multilingual · speech-translation · language-identification · voice-activity-detection · audio · pytorch · openai · whisper · cli · inference · seq2seq · transformer | OpenAI's reference implementation and weights for Whisper, a Transformer speech model that transcribes and translates audio, plus language ID. Ships a CLI and Python API across six model sizes (tiny to large-v3, turbo). | View Project → |
| pi | ★ 109K | AI Coding | TypeScript | coding-agent · cli · tui · agent-runtime · unified-llm-api · tool-calling · monorepo · typescript · self-extensible · terminal · sandboxing · containerization · model-catalog · session-management · state-management | Pi is a TypeScript monorepo agent harness: an interactive coding-agent CLI, an agent runtime with tool calling/state, a unified multi-provider LLM API, and a differential-rendering TUI. MIT licensed and self-extensible via custom providers and sandbox extensions. | View Project → |
| TradingAgents | ★ 108K | AI Agents | Python | multi-agent · finance · trading · langgraph · llm · research · backtesting · risk-management · sentiment-analysis · technical-analysis · cli · docker · ollama · multi-provider | TradingAgents is a LangGraph-based multi-agent framework that simulates a trading firm: analyst, researcher, trader, risk and portfolio agents debate market data and emit trading decisions. Supports many LLM providers and local Ollama, with CLI, Docker and backtesting. | View Project → |
| caveman | ★ 108K | Skills | Go | agent skill · token reduction · prompt engineering · coding agents · output compression · input compression · proxy · middleware · Claude Code · Codex | A cross-agent skill, local proxy, and middleware that reduce token use by compressing agent prose and tool results while preserving originals for retrieval. Supports 30+ coding agents and integrations including LangChain and Vercel AI SDK. | View Project → |
| gemini-cli | ★ 107K | AI Coding | TypeScript | cli · terminal-agent · gemini · mcp-client · ai-agent · code-generation · sandboxing · headless-mode · google-search-grounding · typescript · apache-2.0 · context-files · checkpointing · vscode-companion | Google's open-source AI coding agent for the terminal, built on Gemini with 1M-token context. Supports file/shell/web tools, MCP servers, sandboxing, headless scripting and a VS Code companion. | View Project → |
| open-design | ★ 98K | AI Tools | TypeScript | design-generation · coding-agents · agent-skills · mcp · byok · local-first · desktop-app · prototyping · design-systems · ui-generator · image-generation · video-generation · claude-code · codex · cursor · figma-alternative | Local-first desktop design studio that turns an installed coding-agent CLI (Claude Code, Codex, Cursor, DeepSeek Harness, 20+ others) into a design engine. It composes reusable skills and DESIGN.md design systems to produce prototypes, decks, dashboards, images and video, sandboxed-previewed and exported as HTML/PDF/PP | View Project → |
| agent-skills | ★ 98K | Skills | JavaScript | agent-skills · claude-code · cursor · codex · copilot · slash-commands · workflow · tdd · code-review · spec-driven · context-engineering · subagents · hooks · evals | MIT-licensed pack of 25 Markdown engineering skills plus 9 slash commands that make AI coding agents follow spec, plan, build, test, review and ship workflows across Claude Code, Cursor, Codex, Copilot and more. | View Project → |
| awesome-mcp-servers | ★ 95K | MCP | — | mcp · awesome-list · curated-list · mcp-servers · model-context-protocol · directory · resources · ai · developer-tools · discovery | Curated awesome-list of Model Context Protocol (MCP) servers, organized by category with links and descriptions. Useful as a discovery index for finding MCP servers to plug into LLM clients and agents; contains no code itself. | View Project → |
| vllm | ★ 92K | AI Frameworks | Python | inference-engine · llm-serving · paged-attention · continuous-batching · quantization · speculative-decoding · moe · openai-compatible-api · kv-cache · distributed-inference · lora · structured-output · cuda · rocm · tpu · throughput | vLLM is a high-throughput, memory-efficient LLM inference and serving engine built on PagedAttention, continuous batching, and optimized CUDA/ROCm kernels. It exposes an OpenAI-compatible server with quantization, LoRA, speculative decoding, and disaggregated serving across 200+ model architectures. | View Project → |
| ragflow | ★ 91K | RAG | Go | rag · retrieval-augmented-generation · agentic-retrieval · context-engine · deep-document-understanding · document-parsing · knowledge-base · grounded-citations · hybrid-search · reranking · embeddings · agentic-workflow · mcp · ingestion-pipeline · multi-modal · self-hosted | RAGFlow is an open-source RAG engine combining deep document understanding with agentic retrieval and workflow orchestration. It chunks heterogeneous files (PDF, Office, scans), supports configurable LLMs/embeddings, MCP, and produces grounded answers with traceable citations. Self-hostable via Docker. | View Project → |
| servers | ★ 91K | MCP | TypeScript | mcp · mcp-servers · model-context-protocol · reference-implementations · llm-tools · filesystem · git · memory · sequential-thinking · fetch · anthropic · typescript · nodejs · stdio · developer-tools | Official collection of reference MCP servers (filesystem, git, fetch, memory, sequential-thinking, time) plus a catalog of community servers. Use it as educational examples and copy-paste starting points for giving LLMs secure, controlled access to tools and data. | View Project → |
| taste-skill | ★ 90K | Skills | JavaScript | agent-skills · frontend · design · ui · claude-code · codex · cursor · anti-slop · prompt-engineering · image-generation · gsap · typography · layout · motion · vibecoding · skill-pack | Taste Skill is a portable Agent Skills pack that pushes AI coding agents toward well-designed frontends instead of generic template output. It ships code-emitting design skills (layout, typography, motion, redesign audits) plus image-generation skills for web, mobile, and brand-kit reference boards, installable via npx | View Project → |
| NextChat | ★ 89K | AI Tools | TypeScript | chat-ui · llm · multi-model · cross-platform · self-hosted · nextjs · tauri · pwa · mcp · prompt-templates · streaming · markdown · i18n · desktop-app · privacy-first · plugins | NextChat is a lightweight, cross-platform chat UI (web PWA, desktop via Tauri) for OpenAI, Claude, Gemini, DeepSeek and 100+ model APIs. It supports prompt templates, plugins, MCP tools, artifacts and streaming Markdown, and can be self-hosted on Vercel with local-only chat storage. | View Project → |
| OpenHands | ★ 89K | AI Coding | TypeScript | coding-agents · self-hosted · agent-canvas · acp · claude-code · codex · automations · webhooks · slack · github · docker-sandbox · multi-backend · electron · cli · control-center · developer-tools | OpenHands Agent Canvas is a self-hosted control center UI for running coding agents (OpenHands, Claude Code, Codex, Gemini, any ACP agent) across local, Docker, VM, and cloud backends, with scheduled and event-driven automations wired into Slack, GitHub, and Linear. | View Project → |
| OpenHands | ★ 89K | AI Coding | TypeScript | Python · Coding Agent · OpenHands | Open platform for software agents that write, run, and debug code like a developer. | View Project → |
| odysseus | ★ 88K | AI Tools | Python | self-hosted · ai-workspace · chat-ui · agents · mcp · rag · local-models · deep-research · memory · skills · documents · email · calendar · notes · image-generation · docker | Odysseus is a self-hosted AI workspace bundling chat, agents, MCP servers, RAG memory, deep research, model serving and productivity apps (docs, email, notes, calendar) behind one Docker-deployed FastAPI UI. Best for developers wanting a private, local-model-first alternative to hosted AI chat platforms. | View Project → |
| worldmonitor | ★ 87K | MCP | TypeScript | mcp · mcp-server · osint · geopolitics · news-aggregation · situational-awareness · monitoring · real-time · agent-skills · local-ai · ollama · dashboard · tauri · cli · sdk · intelligence | World Monitor is a real-time OSINT and geopolitical intelligence dashboard that AI-synthesizes news feeds, instability scores, and infrastructure signals. It ships an MCP server, agent-skills manifest, REST API, CLI, and multi-language SDKs, and can run entirely on local Ollama models. | View Project → |
| Agent-Reach | ★ 84K | AI Tools | Python | ai-agent · agent-infrastructure · cli · mcp · web-scraping · search · twitter · reddit · youtube-transcript · bilibili · xiaohongshu · rss · cookie-auth · skills · no-api-key · python | Agent Reach is a Python CLI capability layer that gives AI agents internet reading/search across Twitter, Reddit, YouTube, GitHub, Bilibili, Xiaohongshu, RSS and more. It routes each platform to a working free backend (no paid APIs), registers an MCP/Exa search channel and SKILL.md, and ships a doctor command to check | View Project → |
| Understand-Anything | ★ 84K | Skills | TypeScript | knowledge-graph · codebase-analysis · code-understanding · agent-skills · claude-code-plugin · multi-agent · static-analysis · tree-sitter · semantic-search · code-onboarding · developer-tools · interactive-dashboard · codex · cursor · gemini-cli · llm-wiki | Multi-platform AI coding skill that runs an LLM + tree-sitter multi-agent pipeline over a repo to build an interactive knowledge graph with search, guided tours, diff-impact analysis, and codebase Q&A. | View Project → |
| crawl4ai | ★ 84K | AI Tools | Python | web-scraping · web-crawler · llm · markdown · rag · data-extraction · structured-extraction · deep-crawl · browser-automation · async · docker · api-server · chunking · playwright · anti-bot · caching | Async Python web crawler that converts pages into clean, LLM-ready Markdown and structured JSON via CSS/XPath or LLM extraction. Includes deep-crawl, caching, sessions/proxies, and a Docker FastAPI server, aimed at RAG pipelines and agent data ingestion. | View Project → |
| paperclip | ★ 83K | AI Agents | TypeScript | multi-agent orchestration · agent management · agent workflows · task management · agent governance · human-in-the-loop · cost controls · scheduled agents · audit logs · multi-organization | An open-source control plane for coordinating AI agents as a team: assign goals and tasks, schedule runs, review work, manage permissions, and track budgets. Works with agents and runtimes including OpenClaw, Claude Code, Codex, and Cursor. | View Project → |
| deer-flow | ★ 83K | AI Agents | Python | multi-agent · superagent · deep-research · subagents · sandbox · long-term-memory · skills · langgraph · harness · workflow-orchestration · coding-agent · podcast-generation · report-generation · message-gateway · human-in-the-loop · self-hosted | ByteDance's DeerFlow 2.0 is a SuperAgent harness that orchestrates subagents, sandboxes, memory, skills and tools to run research, coding and content tasks lasting minutes to hours. Ships with a web UI, TUI, IM gateway, MCP server and LangGraph backend. | View Project → |
| lobehub | ★ 83K | AI Agents | TypeScript | multi-agent · agent-collaboration · agent-teams · agent-builder · mcp · skills · self-hosted · chat-ui · knowledge-base · scheduling · memory · workspace · plugins · llm-gateway · typescript · nextjs | LobeHub is a self-hostable multi-agent workspace that treats agents as the unit of work: it hires, schedules, and reports on an AI team. Includes an agent builder, 10k+ skills and MCP plugin library, shared pages/workspaces, and editable personal memory. | View Project → |
| lobe-chat | ★ 83K | AI Tools | TypeScript | TypeScript · Chat · Agents | Extensible open-source AI chat framework with agents, plugins, and modern UX. | View Project → |
| Prompt-Engineering-Guide | ★ 79K | Skills | MDX | prompt-engineering · context-engineering · llm · rag · agents · chain-of-thought · few-shot · react-prompting · tree-of-thoughts · adversarial-prompting · evaluation · education · resources · generative-ai · openai · chatgpt | Maintained guide and course site from DAIR.AI covering prompt engineering, context engineering, RAG and AI agents. Best used as a reference/learning resource for prompting techniques, model notes, papers and notebooks rather than a runnable library. | View Project → |
| learn-claude-code | ★ 77K | AI Agents | Python | agent-harness · tutorial · educational · claude-code · agent-loop · tool-use · subagents · context-compaction · permissions · hooks · memory · task-system · agent-teams · mcp · python · from-scratch | Educational Python tutorial that builds a nano Claude Code-like agent harness from zero, with 17 lessons covering agent loops, tool use, permissions, hooks, subagents, context compaction, memory, MCP plugins and agent teams. | View Project → |
| unsloth | ★ 77K | AI Frameworks | Python | fine-tuning · llm · lora · qlora · grpo · reinforcement-learning · gguf · quantization · inference · local-llm · diffusion · tts · training · desktop-app · openai-api · mcp | Unsloth is a Python training and inference stack plus desktop/web UI for running, fine-tuning and RL-training LLMs and diffusion models 2x faster with less VRAM. Supports LoRA/QLoRA/full fine-tuning, GGUF/MLX export, local OpenAI-compatible serving, RAG and agent hooks. Apache-2.0. | View Project → |
| openai-cookbook | ★ 76K | AI Tools | Jupyter Notebook | openai-api · cookbook · examples · jupyter-notebooks · tutorials · rag · embeddings · fine-tuning · function-calling · agents · prompt-engineering · evals · guardrails · multimodal · gpt-4 · best-practices | Official OpenAI repo of runnable Jupyter notebooks and articles showing how to use the OpenAI API: RAG, embeddings, fine-tuning, function calling, agents, image generation, evals and prompt techniques. Best as a copy-paste reference for developers building LLM apps. | View Project → |
| ai-agents-for-beginners | ★ 75K | AI Agents | Jupyter Notebook | ai-agents · education · course · microsoft-agent-framework · semantic-kernel · autogen · multi-agent · agentic-rag · mcp · a2a · agent-memory · context-engineering · browser-use · foundry-local · production-agents · jupyter-notebook | Microsoft's 18-lesson course for building AI agents, with Jupyter notebooks using Microsoft Agent Framework, Foundry, Semantic Kernel and AutoGen. Covers tool use, agentic RAG, multi-agent and planning patterns, memory, protocols (MCP/A2A), local agents and production deployment. | View Project → |
| orca | ★ 75K | AI Coding | TypeScript | agent-ide · parallel-agents · agent-orchestration · git-worktrees · terminal · claude-code · codex · opencode · computer-use · mobile-companion · ssh-remote · code-review · cli · electron · workflow-automation · multi-agent | Orca is a desktop/mobile Agent Development Environment for running many coding agents (Claude Code, Codex, OpenCode, Cursor, etc.) in parallel git worktrees under your own subscription. It adds terminals, diff review, GitHub/Linear integration, SSH remote runtimes, a CLI and computer-use control. | View Project → |
| MiroFish | ★ 74K | AI Agents | Python | multi-agent-simulation · swarm-intelligence · agent-memory · knowledge-graph · graphrag · prediction · social-simulation · digital-twin · llm · python · zep · public-opinion-analysis · financial-forecasting · persona-generation · simulation-engine · openai-compatible | MiroFish is a multi-agent simulation engine that turns seed material (news, policy drafts, financial signals) into a digital world of thousands of memory-equipped agents, then simulates futures and returns prediction reports. Built on OASIS/CAMEL with GraphRAG and Zep memory; supports OpenAI-compatible LLMs. | View Project → |
| headroom | ★ 74K | AI Tools | Python | LLM context compression · token optimization · AI agents · MCP server · RAG · JSON compression · code compression · LLM proxy · cross-agent memory | Python and TypeScript library, local proxy, and MCP server that compress tool outputs, logs, files, and RAG chunks before they reach an LLM. Includes coding-agent integrations and local retrieval of original content. | View Project → |
| stable-diffusion | ★ 73K | AI Tools | Jupyter Notebook | image-generation · text-to-image · image-to-image · latent-diffusion · inpainting · open-source-model · image-synthesis | Reference code and checkpoints for Stable Diffusion v1, a latent text-to-image diffusion model conditioned by CLIP. Includes scripts for text-to-image generation, image-to-image editing, and inpainting, with PyTorch and Diffusers support. | View Project → |
| ruflo | ★ 73K | AI Agents | TypeScript | multi-agent · swarm-intelligence · agent-orchestration · agent-harness · mcp-server · claude-code · codex · persistent-memory · self-learning · federation · rag · vector-memory · skills · plugins · autonomous-agents · typescript | Ruflo (formerly claude-flow) is an agent meta-harness for Claude Code and Codex: npx ruflo init adds 100+ specialized agents, coordinated swarms, self-learning vector memory, RAG, federation, MCP tools, plugins and hooks on top of your existing coding agent. | View Project → |
| claude-flow | ★ 73K | AI Agents | TypeScript | JavaScript · Claude · Swarm | Orchestration layer for swarming Claude-style agents across complex development work. | View Project → |
| career-ops | ★ 73K | AI Agents | JavaScript | job-search · ai-agent · resume · cv · cover-letter · ats · job-tracker · claude-code · local-first · cli · career · interview-prep · automation · ai-job-search · job-application · hiring | Open-source AI job search agent that runs locally in your AI coding CLI. Scans job portals, evaluates listings into an A-H report with a 1-5 score, tailors CVs and cover letters, and tracks applications. Local-first, human-in-the-loop. | View Project → |
| codegraph | ★ 72K | AI Coding | C | code-graph · code-intelligence · knowledge-graph · static-analysis · tree-sitter · mcp-server · local-first · context-engineering · code-indexing · multi-language · dead-code-detection · auto-sync · cli · agent-context | CodeGraph builds a local, pre-indexed code knowledge graph and exposes it to coding agents (Claude Code, Cursor, Codex, Copilot, Gemini and more) via an MCP server, auto-syncing on file changes. It returns surgical cross-file context to cut tokens and tool calls, and runs 100% locally with no cloud dependency. | View Project → |
| daytona | ★ 72K | AI Tools | — | AI code execution · agent runtime · sandbox · code interpreter · computer use · MCP · isolated environments · persistent state | Daytona provides isolated, persistent sandboxes and SDKs for executing AI-generated code and agent workflows, with process, filesystem, and computer-use tools. The repository states that it is no longer maintained as of June 2026. | View Project → |
| MetaGPT | ★ 71K | AI Agents | Python | multi-agent · llm · agent-framework · software-engineering · sop · natural-language-programming · agent-orchestration · code-generation · data-interpreter · role-based-agents · autonomous-agents · workflow-automation · python · openai | MetaGPT is a multi-agent framework that assigns roles like product manager, architect, and engineer to LLMs, turning a one-line requirement into a full software project through SOPs. It also ships a Data Interpreter agent for data analysis and code generation. | View Project → |
| MetaGPT | ★ 71K | AI Agents | Python | Python · MetaGPT · Software Agent | Software-company-in-a-box: multi-agent collaboration that writes software from a one-line brief. | View Project → |
| impeccable | ★ 70K | Skills | JavaScript | design · skills · ai-coding-agents · frontend · ux · anti-patterns · detector-rules · accessibility · css · html · cli · browser-extension · live-iteration · design-system · code-review · prompt-pack | Impeccable is a design skill pack for AI coding agents: 1 skill, 24 commands (/impeccable audit, polish, critique, animate), plus 61 deterministic detector rules and a browser extension that flag generic AI frontend design. Installs into Claude Code, Cursor, Codex, Copilot and more; the CLI rules run without an LLM. | View Project → |
| OpenSpec | ★ 70K | AI Coding | TypeScript | spec-driven-development · sdd · ai-coding · context-engineering · slash-commands · cli · planning · prd · agent-skills · sdlc · brownfield · markdown · developer-tools · workflow · typescript · npm | OpenSpec is a Node CLI plus slash-command workflow that adds a Markdown spec layer to AI coding assistants: you propose a change, review generated proposal/specs/design/tasks, then let the agent implement and archive it. Works across 30+ tools and brownfield repos. | View Project → |
| archify | ★ 69K | Skills | JavaScript | agent-skill · claude-skill · coding-agents · diagrams · diagram-as-code · architecture-diagram · sequence-diagram · data-flow-diagram · code-visualization · mermaid-alternative · text-to-diagram · system-design · json-ir · svg-export · developer-tools · cli | Archify is an agent skill for Cursor, Claude Code, Codex CLI, and OpenCode that turns a repo or description into validated architecture, workflow, sequence, data-flow, and lifecycle diagrams. Agents emit typed JSON IR that compiles deterministically to self-contained interactive HTML/SVG with exports and delta review. | View Project → |
| cline | ★ 69K | AI Coding | TypeScript | autonomous-agent · coding-agent · ide-extension · cli · vscode · jetbrains · sdk · mcp · multi-agent · computer-use · headless-ci · scheduled-tasks · human-in-the-loop · plan-and-act · checkpoints · typescript | Cline is an open source autonomous coding agent available as a VS Code/JetBrains extension, terminal CLI, desktop app, and Node SDK. It plans and executes edits, runs shell commands, supports MCP servers, multi-agent teams, scheduled runs, and any model provider. | View Project → |
| OmniRoute | ★ 69K | AI Tools | TypeScript | ai-gateway · llm-router · openai-proxy · multi-provider · auto-fallback · token-compression · mcp · a2a · quota-management · free-tier · claude-code · cursor · codex · cline · copilot · self-hosted | OmniRoute is an MIT-licensed AI gateway that unifies 350+ LLM providers behind one OpenAI-compatible endpoint with quota-aware auto-fallback, token compression, and MCP/A2A support. It plugs into coding agents like Claude Code, Cursor, and Cline and self-hosts as CLI, Docker, desktop, or PWA. | View Project → |
| openinterpreter | ★ 68K | AI Coding | Rust | coding-agent · terminal · open-models · Rust · ACP · MCP · sandboxing · computer-use · skills · harnesses | A Rust terminal coding agent that emulates multiple agent harnesses for low-cost and open models. It supports sandboxed command execution, ACP and MCP, shared agent skills, and computer-use testing. | View Project → |
| open-interpreter | ★ 68K | AI Agents | Rust | Python · Interpreter · Local Agent | Local-first agent that runs code on your machine to complete computer tasks. | View Project → |
| docling | ★ 68K | RAG | Python | document-parsing · pdf · ocr · markdown · table-extraction · document-conversion · docling-document · vlm · asr · docx · pptx · xlsx · html · json-export · chunking · rag-ingestion | Docling is a Python library and CLI that parses PDFs, Office files, HTML, EPUB, images and audio into a unified DoclingDocument, exporting to Markdown, HTML or lossless JSON. It is widely used as the document ingestion stage for RAG and agentic AI pipelines. | View Project → |
| annotated_deep_learning_paper_implementations | ★ 67K | AI Frameworks | Python | transformers · pytorch · deep-learning · attention · llm · diffusion · gan · reinforcement-learning · optimizers · lora · education · tutorial · neural-networks · literate-programming · vit · stable-diffusion | LabML's annotated deep learning library: 60+ readable PyTorch implementations of papers with side-by-side notes, covering transformers (GPT, ViT, XL), LoRA, diffusion/Stable Diffusion, GANs, optimizers, and RL. Best as a learning and reference resource rather than a production framework. | View Project → |
| anything-llm | ★ 66K | RAG | JavaScript | rag · local-first · self-hosted · ai-agents · vector-database · document-chat · multi-user · mcp · llm-providers · multimodal · no-code-agent-builder · desktop-app · docker-deployment · openai-compatible · citations · privacy | Self-hosted, local-first AI app: ingest documents into a vector DB and chat with them privately, with built-in agents, no-code agent flows, MCP support, multi-user permissions and a wide range of local or cloud LLMs and embedders. | View Project → |
| claude-code-best-practice | ★ 66K | AI Coding | HTML | Claude Code · best practices · agentic coding · agent teams · subagents · skills · commands · MCP · workflows · context engineering | A practical guide and example repository for Claude Code, covering setup, commands, skills, subagents, MCP, memory, and agent orchestration. Includes implementation examples and workflow documentation for developers adopting agentic coding. | View Project → |
| mem0 | ★ 66K | AI Agents | Python | agent-memory · long-term-memory · memory-layer · vector-store · embeddings · personalization · hybrid-search · temporal-reasoning · entity-linking · state-management · llm · python · typescript-sdk · semantic-search · cli · self-hosted | Mem0 is a drop-in memory layer for AI agents and LLM apps: it extracts, stores and retrieves user, session and agent memories on top of pluggable vector stores. Ships a Python/TS SDK, self-hosted server, CLI and agent skills, with hybrid semantic+keyword+entity retrieval. | View Project → |
| daily_stock_analysis | ★ 66K | AI Agents | Python | stock-analysis · AI-agent · quantitative-finance · LLM · multi-market · financial-news · trading-signals · decision-dashboard · automated-reports · market-data | A Python-based AI stock analysis system that combines multi-market price data, financial news, and LLMs to generate decision dashboards and reports. Includes a conversational stock-analysis agent, web interface, scheduled runs, and notifications. | View Project → |
| warp | ★ 65K | AI Coding | Rust | terminal · coding-agent · agentic-dev-environment · cli-agents · multi-agent · rust · mcp · computer-use · shell · bash · zsh · wasm · developer-productivity · code-review · issue-triage · open-source | Warp is an agentic development environment built on the terminal, written in Rust. It ships a built-in coding agent, lets you plug in external CLI agents like Claude Code, Codex and Gemini CLI, and powers agent 'Factories' that triage issues, write specs, implement changes and review PRs. | View Project → |
| strix | ★ 65K | AI Agents | Python | pentesting · security · multi-agent · red-teaming · vulnerability-scanning · offensive-security · cli · docker · owasp-top-10 · bug-bounty · llm-security · autonomous-agents · ci-cd · sast · dast · exploit-validation | Strix is an open-source AI pentesting agent that autonomously runs your app, finds vulnerabilities, and validates them with working proofs-of-concept. Multi-agent orchestration covers recon, exploitation, and reporting through a developer-first CLI, Docker sandbox, and CI integration. | View Project → |
| nanoGPT | ★ 63K | AI Frameworks | Python | gpt · llm-training · finetuning · transformer · pytorch · gpt-2 · text-generation · minimal · educational · ddp · character-level · openwebtext · shakespeare · reproduction · deprecated · nanochat | Minimal, readable ~300-line PyTorch repo for training and finetuning GPT-2-scale models; reproduces GPT-2 124M on OpenWebText. Author marks it deprecated in favor of the newer nanochat, but it remains a compact hacking/education base. | View Project → |
| last30days-skill | ★ 63K | Skills | Python | agent-skill · deep-research · multi-source-search · recency · trends · reddit · twitter-x · youtube · hacker-news · polymarket · web-search · social-signals · summarization · claude-code · codex · agentskills | An agent skill that searches Reddit, X, YouTube, TikTok, HN, Polymarket, GitHub and the web in parallel, scores results by real engagement, and has an AI agent synthesize a cited last-30-days brief. Installs into Claude Code, Codex, Cursor and 50+ Agent Skills hosts. | View Project → |
| TrendRadar | ★ 62K | AI Tools | Python | news-aggregation · trend-monitoring · public-opinion · rss · mcp-server · push-notifications · llm · ai-summarization · ai-translation · sentiment-analysis · keyword-filtering · docker · github-actions · self-hosted · webhook · python | Python trend/opinion monitor that aggregates multi-platform hot lists and RSS, using LLMs for keyword filtering, translation and daily briefs pushed to WeChat, Feishu, DingTalk, Telegram, email, ntfy, Bark and Slack. Also ships an MCP server for conversational news analysis. | View Project → |
| context7 | ★ 62K | MCP | TypeScript | mcp · mcp-server · documentation · context · llm · vibe-coding · code-docs · developer-tools · typescript · cursor · claude-code · cli · vercel-ai-sdk · api-docs · upstash | Context7 is an MCP server and CLI that pulls version-specific, up-to-date library documentation and code examples directly into LLM prompts. It exposes resolve-library-id and query-docs tools (or ctx7 library/docs commands) so coding agents stop hallucinating outdated APIs. | View Project → |
| OpenMontage | ★ 61K | AI Agents | Python | video production · AI agents · video generation · image generation · text-to-speech · video editing · production pipelines · agent skills · storyboarding · FFmpeg · Remotion | An agentic video production system that coordinates research, scripting, asset generation, editing, and rendering through AI coding assistants. Includes production pipelines, agent skills, a local live storyboard, and integrations with media-generation providers. | View Project → |
| autogen | ★ 61K | AI Agents | Python | multi-agent · agentic-ai · orchestration · llm-agent · mcp · code-execution · human-in-the-loop · event-driven · distributed-runtime · python · dotnet · no-code-gui · benchmarking · openai · autogen · maintenance-mode | Microsoft's multi-agent framework for agentic AI apps, with layered Core/AgentChat/Extensions APIs, MCP tool support, a no-code Studio GUI, and AutoGenBench evals. Now in maintenance mode; new projects are directed to Microsoft Agent Framework. | View Project → |
| litellm | ★ 59K | AI Frameworks | Python | ai-gateway · llm-gateway · llmops · openai-compatible · proxy-server · cost-tracking · guardrails · load-balancing · mcp-gateway · a2a · multi-provider · python-sdk · rust · observability · virtual-keys · rate-limiting | LiteLLM is an open-source AI gateway and Python SDK that calls 100+ LLM providers (OpenAI, Anthropic, Bedrock, Vertex, vLLM, Ollama) in OpenAI format. The proxy adds virtual keys, spend tracking, guardrails, load balancing, logging, plus MCP tool and A2A agent routing. | View Project → |
Vibe Coding Discover