Report Date: 2026-04-06
Analysis Window: 30 days
Languages Tracked: python, typescript, rust, go, javascript
Top N: 50
Overview
过去30天，GitHub AI 开源趋势由 'Agent' 类别主导，占据了过半的热门项目。开发者们高度关注 Agent 的工具链和基础设施，特别是与 Claude 相关的项目， 显示出构建更强大、更专业化 AI 系统的趋势。语言方面，Python 仍是主流，但 Rust 和 Go 在性能关键型组件中的应用也日益增多。
Top 50 Repositories
| Rank | Repository | Category | Language | Heat_7 | Heat_30 | Score |
|---|---|---|---|---|---|---|
| 1 | everything-claude-code | infra | javascript | 24,270 | 79,260 | 1,304 |
| 2 | hermes-agent | agent | python | 9,940 | 24,111 | 1,256 |
| 3 | OpenViking | agent | python | 0 | 16,523 | 55 |
| 4 | TradingAgents | agent | python | 0 | 16,280 | 54 |
| 5 | impeccable | javascript | 0 | 15,660 | 52 | |
| 6 | last30days-skill | agent | python | 0 | 15,180 | 50 |
| 7 | VibeVoice | python | 10,543 | 12,995 | 495 | |
| 8 | claude-hud | agent | javascript | 0 | 12,943 | 43 |
| 9 | GitNexus | agent | typescript | 0 | 12,538 | 41 |
| 10 | ruflo | agent | typescript | 0 | 11,168 | 37 |
| 11 | codex | agent | rust | 5,169 | 10,489 | 256 |
| 12 | OpenSpec | typescript | 0 | 9,831 | 32 | |
| 13 | deepagents | agent | python | 0 | 9,451 | 31 |
| 14 | promptfoo | llm | typescript | 0 | 8,723 | 29 |
| 15 | BitNet | llm | python | 0 | 8,697 | 28 |
| 16 | marketingskills | agent | javascript | 0 | 7,855 | 26 |
| 17 | onyx | llm | python | 4,741 | 7,162 | 825 |
| 18 | claude-skills | llm | python | 0 | 7,083 | 23 |
| 19 | notebooklm-py | agent | python | 0 | 6,303 | 21 |
| 20 | chrome-devtools-mcp | agent | typescript | 0 | 5,862 | 19 |
| 21 | new-api | llm | go | 0 | 5,858 | 19 |
| 22 | pentagi | agent | go | 0 | 5,161 | 17 |
| 23 | hindsight | agent | python | 0 | 5,094 | 16 |
| 24 | supermemory | typescript | 0 | 4,542 | 15 | |
| 25 | ai-hedge-fund | python | 0 | 4,121 | 13 | |
| 26 | open-swe | agent | python | 0 | 3,966 | 13 |
| 27 | fff.nvim | agent | rust | 2,120 | 2,679 | 145 |
| 28 | openclaw-mission-control | agent | typescript | 0 | 2,007 | 6 |
| 29 | LocalAI | vision | go | 0 | 1,781 | 31 |
| 30 | one-api | llm | javascript | 0 | 1,507 | 5 |
| 31 | forgecode | llm | rust | 663 | 1,306 | 32 |
| 32 | 9router | llm | javascript | 406 | 1,129 | 21 |
| 33 | screenpipe | agent | rust | 414 | 1,000 | 21 |
| 34 | nono | agent | rust | 0 | 928 | 3 |
| 35 | bifrost | go | 216 | 834 | 12 | |
| 36 | docker-agent | agent | go | 0 | 705 | 2 |
| 37 | claude-squad | agent | go | 0 | 654 | 2 |
| 38 | cocoindex | rust | 0 | 475 | 1 | |
| 39 | claude-code-skills | infra | javascript | 0 | 167 | 0 |
| 40 | goose | agent | rust | 2,800 | 0 | 649 |
| 41 | claude-howto | agent | python | 15,128 | 0 | 648 |
| 42 | agent-browser | agent | rust | 0 | 0 | 228 |
| 43 | memvid | rag | rust | 0 | 0 | 225 |
| 44 | pi-mono | agent | typescript | 0 | 0 | 213 |
| 45 | agent-framework | agent | python | 608 | 0 | 192 |
| 46 | oh-my-openagent | agent | typescript | 4,031 | 0 | 172 |
| 47 | DeepScientist | typescript | 0 | 0 | 151 | |
| 48 | TaxHacker | llm | typescript | 2,204 | 0 | 94 |
| 49 | RAG-Anything | rag | python | 500 | 0 | 80 |
| 50 | PaddleOCR | llm | python | 1,736 | 0 | 74 |
Category Breakdown
llm (10 projects)
仍然是核心领域，有10个项目，主要集中在模型管理、测试（如 promptfoo）和特定场景应用（如 TaxHacker）。
Average Heat_7: 975
Average Score: 115
- promptfoo - Test your prompts, agents, and RAGs. Red teaming/pentesting/vulnerability scanning for AI. Compare performance of GPT, Claude, Gemini, Llama, and more. Simple declarative configs with command line ...
- BitNet - Official inference framework for 1-bit LLMs
- onyx - Open Source AI Platform - AI Chat with advanced features that works with every LLM
- claude-skills - 220+ Claude Code skills & agent plugins for Claude Code, Codex, Gemini CLI, Cursor, and 8 more coding agents — engineering, marketing, product, compliance, C-level advisory.
- new-api - A unified AI model hub for aggregation & distribution. It supports cross-converting various LLMs into OpenAI-compatible, Claude-compatible, or Gemini-compatible formats. A centralized gateway for p...
- one-api - LLM API 管理 & 分发系统，支持 OpenAI、Azure、Anthropic Claude、Google Gemini、DeepSeek、字节豆包、ChatGLM、文心一言、讯飞星火、通义千问、360 智脑、腾讯混元...
- forgecode - AI enabled pair programmer for Claude, GPT, O Series, Grok, Deepseek, Gemini and 300+ models
- 9router - Connect All AI Code Tools (Claude Code, Cursor, Antigravity, Copilot, Codex, Gemini, OpenCode, Cline, OpenClaw...) to 40+ AI Providers & 100+ Models
- TaxHacker - Self-hosted AI accounting app. LLM analyzer for receipts, invoices, transactions with custom prompts and categories
- PaddleOCR - Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit that bridges the gap between images/PDFs and LLMs. Supports 100+ languages.
vision (1 projects)
仅有一个项目（LocalAI）上榜，表明当前趋势热点更多地集中在语言模型和 Agent 系统上，而非计算机视觉。
Average Heat_7: 0
Average Score: 31
- LocalAI - LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
rag (2 projects)
一个虽小众但重要的领域。像 memvid 这样的项目正在探索新的方法，例如“无服务器、单文件记忆层”，以简化复杂的 RAG 流程。
Average Heat_7: 250
Average Score: 152
- memvid - Memory layer for AI Agents. Replace complex RAG pipelines with a serverless, single-file memory layer. Give your agents instant retrieval and long-term memory.
- RAG-Anything - "RAG-Anything: All-in-One RAG Framework"
infra (2 projects)
尽管只有两个项目，但该类别的平均热度和评分遥遥领先，这得益于像 everything-claude-code 这样的基础性项目。这凸显了社区对AI开发底层设施的高度重视。
Average Heat_7: 12,135
Average Score: 652
- everything-claude-code - The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
- claude-code-skills - Plugin suite + bundled MCP servers for Claude Code. Full delivery lifecycle: Agile pipeline with multi-model AI review, project bootstrap, documentation generation, codebase audits, performance opt...
agent (27 projects)
最活跃的类别，包含27个项目。重点在于框架、编排和专用技能，表明 Agent 概念正从基础理论走向更成熟的实际应用。
Average Heat_7: 1,489
Average Score: 156
- hermes-agent - The agent that grows with you
- OpenViking - OpenViking is an open-source context database designed specifically for AI Agents(such as openclaw). OpenViking unifies the management of context (memory, resources, and skills) that Agents need th...
- TradingAgents - TradingAgents: Multi-Agents LLM Financial Trading Framework
- last30days-skill - AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary
- claude-hud - A Claude Code plugin that shows what's happening - context usage, active tools, running agents, and todo progress
- GitNexus - GitNexus: The Zero-Server Code Intelligence Engine - GitNexus is a client-side knowledge graph creator that runs entirely in your browser. Drop in a GitHub repo or ZIP file, and get an interactive ...
- ruflo - 🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade arch...
- codex - Lightweight coding agent that runs in your terminal
- deepagents - Agent harness built with LangChain and LangGraph. Equipped with a planning tool, a filesystem backend, and the ability to spawn subagents - well-equipped to handle complex agentic tasks.
- marketingskills - Marketing skills for Claude Code and AI agents. CRO, copywriting, SEO, analytics, and growth engineering.
- notebooklm-py - Unofficial Python API and agentic skill for Google NotebookLM. Full programmatic access to NotebookLM's features—including capabilities the web UI doesn't expose—via Python, CLI, and AI agents ...
- chrome-devtools-mcp - Chrome DevTools for coding agents
- pentagi - Fully autonomous AI Agents system capable of performing complex penetration testing tasks
- hindsight - Hindsight: Agent Memory That Learns
- open-swe - An Open-Source Asynchronous Coding Agent
- fff.nvim - The fastest and the most accurate file search toolkit for AI agents, Neovim, Rust, C, and NodeJS
- openclaw-mission-control - AI Agent Orchestration Dashboard - Manage AI agents, assign tasks, and coordinate multi-agent collaboration via OpenClaw Gateway.
- screenpipe - Run agents that work for you based on what you do. AI finally knows what you are doing
- nono - Kernel-enforced agent sandbox. Capability-based isolation with secure key management, atomic rollback, cryptographic immutable audit chain of provenance. Run your agents in a zero-trust environment.
- docker-agent - AI Agent Builder and Runtime by Docker Engineering
- claude-squad - Manage multiple AI terminal agents like Claude Code, Codex, OpenCode, and Amp.
- goose - an open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM
- claude-howto - A visual, example-driven guide to Claude Code — from basic concepts to advanced agents, with copy-paste templates that bring immediate value.
- agent-browser - Browser automation CLI for AI agents
- pi-mono - AI agent toolkit: coding agent CLI, unified LLM API, TUI & web UI libraries, Slack bot, vLLM pods
- agent-framework - A framework for building, orchestrating and deploying AI agents and multi-agent workflows with support for Python and .NET.
- oh-my-openagent - omo; the best agent harness - previously oh-my-opencode
(8 projects)
Average Heat_7: 1,344
Average Score: 96
- impeccable - The design language that makes your AI harness better at design.
- VibeVoice - Open-Source Frontier Voice AI
- OpenSpec - Spec-driven development (SDD) for AI coding assistants.
- supermemory - Memory engine and app that is extremely fast, scalable. The Memory API for the AI era.
- ai-hedge-fund - An AI Hedge Fund Team
- bifrost - Fastest enterprise AI gateway (50x faster than LiteLLM) with adaptive load balancer, cluster mode, guardrails, 1000+ models support & <100 µs overhead at 5k RPS.
- cocoindex - Data transformation framework for AI. Ultra performant, with incremental processing. 🌟 Star if you like it!
- DeepScientist - Now, Stronger AI Pushes Frontiers, Stronger Our Shared Future.
Dark Horse Projects
“黑马”项目榜单揭示了那些迅速获得巨大关注的新兴项目。affaan-m/everything-claude-code 和 luongnv89/claude-howto 是典型代表，其极高的7日热度值反映了社区的强烈兴趣。这表明，实用的指南和强大的基础设施工具，特别是围绕 Claude 生态的，能够在一夜之间吸引大量关注。
| Repository | Score | Heat_30 | Heat_7 | Category |
|---|---|---|---|---|
| everything-claude-code | 1,304 | 79,260 | 24,270 | infra |
| hermes-agent | 1,256 | 24,111 | 9,940 | agent |
| VibeVoice | 495 | 12,995 | 10,543 | |
| codex | 256 | 10,489 | 5,169 | agent |
| onyx | 825 | 7,162 | 4,741 | llm |
| fff.nvim | 145 | 2,679 | 2,120 | agent |
| goose | 649 | 0 | 2,800 | agent |
| claude-howto | 648 | 0 | 15,128 | agent |
| agent-browser | 228 | 0 | 0 | agent |
| memvid | 225 | 0 | 0 | rag |
| pi-mono | 213 | 0 | 0 | agent |
| agent-framework | 192 | 0 | 608 | agent |
| oh-my-openagent | 172 | 0 | 4,031 | agent |
| DeepScientist | 151 | 0 | 0 |
Consecutive Appearances
像 openai/codex（11周）和 NousResearch/hermes-agent（5周）这样的长期上榜者，标志着这些项目拥有持续的社区动力和用户基础。尽管新项目层出不穷，它们依然在榜单上占有一席之地，证明了它们在 AI Agent 领域的基石地位和活跃的开发状态。
| Repository | Weeks in Top | Category | Current Heat_7 |
|---|---|---|---|
| everything-claude-code | 3 | infra | 24,270 |
| hermes-agent | 5 | agent | 9,940 |
| OpenViking | 5 | agent | 0 |
| TradingAgents | 3 | agent | 0 |
| impeccable | 5 | 0 | |
| last30days-skill | 2 | agent | 0 |
| VibeVoice | 2 | 10,543 | |
| claude-hud | 3 | agent | 0 |
| GitNexus | 7 | agent | 0 |
| ruflo | 3 | agent | 0 |
| codex | 11 | agent | 5,169 |
| deepagents | 4 | agent | 0 |
| promptfoo | 4 | llm | 0 |
| BitNet | 4 | llm | 0 |
| marketingskills | 4 | agent | 0 |
| onyx | 2 | llm | 4,741 |
| claude-skills | 3 | llm | 0 |
| notebooklm-py | 4 | agent | 0 |
| new-api | 2 | llm | 0 |
| supermemory | 2 | 0 | |
| ai-hedge-fund | 5 | 0 | |
| open-swe | 3 | agent | 0 |
| fff.nvim | 2 | agent | 2,120 |
| one-api | 4 | llm | 0 |
| 9router | 3 | llm | 406 |
| docker-agent | 4 | agent | 0 |
| claude-squad | 2 | agent | 0 |
| claude-code-skills | 3 | infra | 0 |
| claude-howto | 2 | agent | 15,128 |
| agent-browser | 4 | agent | 0 |
Highlighted Repositories
affaan-m/everything-claude-code
该项目以极高的分数（1304）和7日热度（24270）高居榜首。作为一个“Agent 性能优化系统”，它本 身不是 Agent，而是用于提升其他 Agent 的元工具。它的流行表明市场对于管理和增强复杂 AI Agent 性能的基础设施需求日益增长，尤其是在 Claude 生态系统中。
microsoft/VibeVoice
这是来自大型科技公司的一个重要新项目。项目描述为“前沿开源语音AI”，其高涨的7日热度（10543）显示了强大的初期市场兴趣。该项目可能预示着在高质量语音合成与交互领域，将迎来新一波开源创新浪潮，而这一领域传统上由闭源模型主导。
luongnv89/claude-howto
这个项目是一个典型的“黑马”。尽管30日热度为零，但其7日热度却飙升至15128。作为一个“可视化、案例驱动的 Claude Code 指南”，它的成功突显了社区对于高质量教程和实用模板的迫切需求，以便更有效地利用像 Claude 这样的高级 AI 工具。
block/goose
这是另一个迅速获得高分（649）的新项目。它使用 Rust 编写，是一个“超越代码建议”的可扩展 AI Agent，能够执行、编辑和测试代码。选择 Rust 语言表明，Agent 工具正在追求更高的性能和安全性，从简单的脚本向更健壮的编译型应用演进。
Methodology
Data Sources:
- GitHub Trending pages
- GitHub API stargazer timestamps
Metrics:
- Heat_7: Stars gained in last 7 days
- Heat_30: Stars gained in last 30 days
- Score: Weighted scoring combining short-term heat and sustained growth
- Formula: 0.6 × stars_1d + 0.3 × (stars_7d / 7) + 0.1 × (stars_30d / 30)
- Emphasizes recent activity (60%) while considering sustained trends (40%)
Filtering:
- Include keywords: %!s(int=30)
- Exclude keywords: %!s(int=30)
- Categories: %!s(int=30)
Ranking:
- Sort by Heat_30 (descending)
- Tie-break by Score (descending)
- Select top %!d(string=ai, llm, agent, gpt, rag, langchain, machine-learning, deep-learning, neural, transformer, embedding, vector)
Limitations:
- Trending data limited to GitHub's trending algorithm
- Star history may be incomplete for repos with >40k stars (API pagination limits)
- LLM-generated commentary is interpretive, not prescriptive
- Weekly snapshots may miss short-lived trends%!(EXTRA string=tutorial, awesome, learning, course, book, string=vision, infra, agent, llm, rag, int=50)
Generated at: 2026-04-06T00:33:21Z