GitHub agents

The agent that learns how you decide - and proves it by never deciding for you. A local-first, loop-native runtime for a team of AI agents on your machine.

RAG-Powered Financial Query Assistant for JPMorgan Chase documents with Hybrid Search, Docker support, and Streamlit UI

Production-ready tutorial for building Retrieval-Augmented Generation (RAG) systems with LangChain.

EduMIND — Multimodal Bilingual Lecture Assistant with Anti-Forget RAG, Whisper ASR, VietMix NMT (CMI-aware), and Human-in-the-Loop Active Learning via Label Studio. Built for code-mixed Vietnamese–English academic environments.

Claude Code 委派 plugin:把任务派给换脑的子 agent(也可 --codex 委派给本机 codex),既省主 Claude 的 token,也借不同模型/harness 的视角。

Cross-agent plugin collection for AI coding CLIs. Delegate tasks and run workflow patterns (council, parallel review, debug) across Claude Code, OpenCode, Gemini CLI, and Codex.

[ICML'26] Codebase for PersistBench: When Should Long-Term Memories Be Forgotten by LLMs to benchmark cross-domain leakage and sycophancy in memory augmented LLMs.

Run the DeepSeek V4 Pro AI model on your desktop for fast coding, research, and problem solving.

A clean, modular, and production-ready WhatsApp chatbot using personal accounts via Baileys.

私有部署的 AI 运维数字员工——自建 RAG 引擎（BM25+向量混合检索+重排序），知识库管理，申告工单全流程，数据不出域。Go + Next.js 全栈，Docker Compose 一键部署。

Free 42-page AI agents crash course v5.0 — 5 days, 3 learning paths, 6 playable games, 13 Anthropic Academy courses, interactive quizzes, XP gamification. From zero to multi-agent systems. MIT licensed.

A factory for building production-grade agent skills — benchmarked against gold standards, autonomously improved via an autoresearch loop, and verified by independent multi-agent consensus.

🧮 CLI for counting tokens in files and directories using tiktoken

A modular AI-powered CLI for Azure Sentinel threat hunting & remediation. Features strict guardrails, cost-aware routing, and automated SOAR workflows (VM isolation, rule creation).

BoxAgnts is an open-source, Rust-based AI Agent ToolBox offering an ultimate out-of-the-box experience. With WebAssembly sandbox, it balances security and flexibility, letting users effortlessly tackle complex tasks as an efficient, trustworthy personal AI assistant.

虚拟信用卡

Swift/SwiftUI library for rendering LLM-generated <artifact> blocks in chat interfaces. Supports Markdown, JSON, CSV, Code, SVG, GeoJSON, HTML, React, Mermaid, LaTeX, Vega-Lite, GLTF, USDZ.

A Claude Skill that builds polished, interactive HTML explainers, with an automatic QA loop and anti-slop visual guards.

ClawTouch MCP server — exposes a real USB-HID keyboard/mouse (Raspberry Pi Pico 2) as Model Context Protocol tools for any LLM agent. MIT.

学习.skill —— 让 Agent 成为你的赛博老师，帮你备课，陪你学习

Open-source Android phone-use agent. Your phone, on autopilot.

The agentic, conversational web browser — talk to Sprig and the open web composes itself into living, cited pages.

Mcp server code to setup muapi to work with clients like Claude, Cursor etc.

口述史录音转写稿 → 忠实整理稿的工具箱（Claude Skill）：去口语化 Pass1 + 对抗式忠实度质量门，存真优先，数据全程本地不外传

OR-Space: a full-lifecycle workspace benchmark for industrial optimization agents

Connect Google ADK (Agent Development Kit) agents to Telegram, Discord, Slack, and WhatsApp with minimal code, featuring real-time cross-device session synchronization.

Token-lean web microfetch for LLM agents: any URL → clean markdown via CLI, MCP server, and Claude Code plugin. Real browser-cookie auth, passkeys, anti-bot reach, on-by-default prompt-injection defense, plus on-device multimodal ASR/OCR. A single Rust binary — not a browser.

Powerline-style statusline for Claude Code — Rust CLI with a TUI configurator, themes and styles

🏛️ 通用型AI认知架构 · 诺亚三代二型 — 抽屉级联上下文管理 / 记忆闭环(Mnemosyne) / 自主进化。实战萃取，MIT开源

Self-hosted, GDPR-safe AI agent for law firms. Zero-cloud, hash-chained audit trail (AI Act art. 12), bring-your-own-model, 9 language editions. Hard fork of willchen96/mike, AGPL-3.0.

裂变创作-Fission Creative — 长篇网文辅助创作系统

Free, browser-based BYOK AI website builder — v0 / Lovable / Bolt.new alternative. Bring your own key from 17 providers, generate & deploy full code, no subscription.

CursorJr — русскоязычный субагент-наставник для Cursor: автоматизация, MCP, Rules, Skills, Canvas и обучение новичков

Personal AI agent — tell it what you want, it figures out how. Runs locally, plans autonomously, writes code, browses the web, and gets things done.

面向智能座舱的云边协同 AI Agent：端侧毫秒级车控，云端声明式 Multi-Agent 与 Skill/DAG 编排，支持 S2S 实时语音、声纹多用户、视觉与 HMI；LLM 只负责理解与规划，VAL 负责确定性安全执行。

Rootsign is an open-source tamper-evident decision and action provenance logging library for AI agents

A runtime-portable 5-agent adjudicator council for universal quality gating of text artifacts.

Hollywood-grade AI video production system: Multi-Agent orchestration, 25-field cinematic shot language, and automated filmmaking pipeline. Built for AI agents.

agent江湖 - 多Agent协作平台

MCP server exposing the full Apple Ads (Search Ads) Campaign Management API v5 — 74 typed tools

Never hit a rate-limit again. Route Claude Code, Cursor & 40+ providers through the subscription you already pay for, cut 40-80% tokens (RTK), stay un-banned. Self-hosted LiteLLM / OpenRouter / 9router alternative — one Go binary, anti-ban + sovereign P2P mesh.

Multi-tenant AI front desk grounded in a tenant's knowledge base via hybrid RAG (vector + BM25 + RRF), with human-inbox escalation

AI argues. Code settles. The losses stay on the page. A real HK + US brokerage account run by agents that must debate every call, settled by code the model never touches. Install the same decision workflow into your own agent: OpenClaw, Claude Code, Codex, or DeepSeek Harness.

Evidence-graded, agent-executable thinking-method skills - 56 frameworks reduced to their working mechanism, honestly graded, and producing a concrete artifact. 56 frameworks + 4 tools + 9 recipes for Claude Code, Codex, and other AI agents. Advanced (Gold) tier.

Free, private skill search for AI agents

The complete Google SEO MCP for Claude Code & Cursor — Search Console + Analytics 4 + 5 cross-platform tools (GSC↔GA4 journey, opportunity matrix, traffic health check, revenue attribution, full landing page diagnosis). Diagnostic intelligence with anti-hallucination provenance, leave-one-out Z-score, real funnels. Read-only by default.

MCP server and local AI control platform for AutoCAD Electrical — Claude, Ollama, web dashboard. macOS + Windows.

Engineering knowledge infrastructure for AI coding agents.

The AI agent harness you can audit: token-waste ledger, leak-proof scoped memory, eval-gated learning, 20+ LLM providers (Claude, OpenAI, Ollama, Grok, Kimi). TypeScript, MIT.

组会skill是一个专为研究生和科研人员打造的skill。它可以帮助你高效地向导师、实验室会议和合作者进行汇报，让你无需每周从头开始撰写重复的更新。

Understand what your AI agents are building. Agent-native, human-sovereign ontology workbench — one markdown graph humans and coding agents read, write, and keep fresh together. Local-first, MCP-native, open source.

Agent control plane for managed OpenShift services

Open-source platform for teams to run AI coding agents in isolated sandboxes (e.g. Codex, Claude Code, OpenCode)

The FULL-VOCAL interaction driven AI system that actively sustains your daily. Simply talk and everything is done.

2026全新ChatGPT Plus充值订阅教程(国内ChatGPT代充购买必看)

Awesome list of 590+ animated pets for the OpenAI Codex CLI — browse, install with npx codex-pet-cli, and submit your own at codex-pet.com

Open-source real-time meeting copilot — live transcription + AI answers in a floating overlay. Anthropic, OpenAI, Gemini, Ollama, or any OpenAI-compatible endpoint.

多模态电商智能导购 Agent - 基于 RAG 的购买决策助手（ByteDance AI Challenge 参赛项目）

Docker image for a self-hosted Docling document parsing server. Converts PDF, DOCX, PPTX, HTML, and more to Markdown/JSON. Powered by IBM Docling. Features sync/async conversion, chunking for RAG, NVIDIA GPU (CUDA) acceleration, optional web UI, offline mode, and persistent model cache. Multi-arch: amd64, arm64.

Apple Human Interface Guidelines archive (1980-2014) - 35 documents optimized for LLM consumption and human exploration. Spanning Lisa, Mac, NeXT, Newton, Aqua, and iOS eras.

Persistent local memory for Claude and ChatGPT. No cloud, no subscription. Self-hosted RAG over MCP.

Agent-agnostic semantic skill for MaxCompute SQL — CLI + bare SKILL.md bundle

GenPark skill: Utilize GenSpark Workspace 4.0 for desktop automation, MS Office integration, and AI meeting notes.

Open-source AI-native host security platform with vulnerability management, compliance auditing, eBPF runtime detection, attack investigation and automated remediation.

102 Claude Code skills across 7 categories -- trading strategies, Azure, VSCode extensions, AI prompts, and custom automation skills

AOI: a model-agnostic perception layer giving computer-use agents adaptive video + audio perception with zero retraining, plus DynaCU-Bench (100 dynamic browser tasks). Paper + benchmark + code.

一个基于 LangChain + FastAPI 的扫地机器人领域智能客服 Agent，支持 RAG 知识库问答、多轮对话、JWT 认证与会话隔离、Redis 缓存与限流。适用于初学者自学RAG和Agent项目

Ultracode making for OpenCode, inspired by Claude Code

A self-growing AI company you run as a desktop app — a governed mesh of agents on a memory that learns how you work. Built on Claude Code.

本体平台（Ontology Platform） 业务规则智能校验平台。基于本体（ontology）治理 + LLM 智能建库 + DSL 规则引擎，为上游 Agent 和业务系统提供规则校验和数据提取能力。

Open-source AI knowledge base & custom chatbot builder — production-ready Next.js SaaS with RAG, document upload, URL scraping, Q&A training, citations, and embeddable chatbot widgets. Stripe, credits, NextAuth, Prisma.

Self-hosted, single-file Indian portfolio tracker with AI insights, Gmail expense tracking

MCP server that lets Claude (or any Model Context Protocol client) drive the CODESYS V3.5 SP22 PLC IDE via its Python ScriptEngine - projects, POUs, build, devices, libraries, and live online control.

Local-first MCP server for Garmin Body Battery, HRV, sleep & training data — Claude/Cursor ready

국가연구개발(정부 R&D) 규정 검토용 MCP 서버 — 혁신법·시행령·행정규칙 55종을 AI로 검색·검토 (Korean national R&D regulations MCP)

Self-hosted AI cinema engine. 5-layer production pipeline with RAG-driven virtual cinematography (1,645 presets), 5 N8N workflows (67 nodes), Remotion compositing, 21-LUT color grading, and 15-model lip-sync. MIT licensed.

鳳問 Ask Audrey — cited AI answers from Audrey Tang's 30-year transcript archive (archive.tw). Cloudflare Workers + Hono; web + LINE bot. 以 AI 檢索唐鳳逐字稿、附出處作答

High-performance Holographic Memory System (HMS) for Node.js, powered by Rust. This library implements Vector Symbolic Architectures (VSA) using Binary Spatter Code (BSC) to enable semantic search, analogical reasoning, and associative memory at scale.

conversation-council (codex/reasonix) — 让 Codex/Reasonix 历史会话以"群聊"参与新讨论。支持 Reasonix v0.53/v1.X + Codex 自动检测。

Vant: Versatile Autonomous Networked Tool - Durable AI memory + cold storage + runtime for agents, with generational evolution, stored with version control. Built for long-running autonomous sessions, using git based persistence.

An intent-based language for agentic systems. Write agents in English. Transpile to async Python.

GitHub Action to sync team AI assets across repositories from sync-config.json

Autonomous prediction market trading bot — Polymarket & Kalshi

Built an AI-powered Resume Analyzer that evaluates resumes against job descriptions, extracts key skills and experience, provides ATS compatibility scores, identifies skill gaps, and generates personalized improvement recommendations using NLP and LLMs.

AI-native NoSQL + vector database with built-in semantic search for RAG, AI agents & LLM memory — the async Python client. Self-hosted Pinecone alternative.

更聚焦普通业务场景的harnass

Skill letting Agents scan a repo, ask what it cannot infer, then generate a complete open source file collection i.e. README, LICENSE, contributing, security, changelog, CI, launch playbook, etc.

Claude Code 增强版，支持本地私有化部署，无云端依赖，支持私有化部署；同时适配DeepSeek、通义千问等国产主流大模型，兼顾OpenAI模型兼容，支持中英双语交互，界面与提示中文适配，且会持续迭代更新，优化性能、适配更多模型，部署简单且功能稳定，在保留原版核心Agent能力的基础上进一步扩展实用功能，提升开发体验。

Device-local HarmonyOS agent phone prototype where the agent becomes the entry point and apps become tools

A modular coding-agent runtime for local models with a CLI, full-screen TUI, VS Code extension, and desktop app built on the same tools, safeguards, and workspace state.

Model Shell Protocol: an application-owned command runtime that gives AI agents OS-like workspace semantics.

GenPark skill: Learn practical AI skills step-by-step through beginner-friendly, habit-forming lessons and structured challenges inspired by Coursiv.io.

A personal financial advisor that runs locally. Plaid CLI + SQLite + Claude Code skills. Fiduciary-grade advice grounded in your real data.

Open-source Android computer-use agent with a self-hostable backend

统一视频剪辑操控 Skill —— 让 AI agent 操控剪映/CapCut 与 Adobe Premiere Pro（支持 CLI/MCP/HTTP）| Unified video editing control skill for AI agents — supports JianYing/CapCut and Adobe Premiere Pro via CLI/MCP/HTTP

A general-purpose interview training officer for tech roles (Claude Code plugin): mock interviews, debriefs, progressive L1-L5 drilling, capability tracking, pipeline management, resume tailoring.

Персональный ИИ-ассистент нового поколения для Windows — голос, зрение, управление ПК, навыки, память.

MCP server that lets your AI assistant read and write Figma in real time — no API token, no rate limits. Works with Claude, Cursor, Copilot, Codex and any MCP client.

通过 HTTP API 让 AI Agent 控制 Android 设备的开源自动化应用

Event-driven AI coding agent for GitHub: watches issues, runs headless Claude Code in isolated git worktrees to triage bugs & features, and opens PRs automatically. Stateless, self-hosted, written in Go.