Assistants & Domain Agents tools for MCP

Collection of production LLM applications with source code across RAG, multi-agent, and tool-use patterns. Good reference for how harness primitives combine in real applications

PrivateGPT is a secure, offline-capable AI tool for querying documents with Large Language Models, offering high-level and low-level APIs for privacy-conscious, context-aware application development github

HKUDS's ultra-lightweight, self-hosted personal AI agent framework: a single readable Python core that combines WebUI/terminal/chat-app surfaces, long-term memory, MCP tools, model routing, multi-agent delegation, and scheduled automation. A practical reference for how a complete personal agent harness can own the full stack without becoming a black-box platform

Official Skyvern browser automation skill — AI-powered browser control using Vision LLMs and computer vision for navigating sites, filling forms, and extracting structured data

短剧、漫剧制作工具 Toonflow：开源 AI 视频创作平台，集成剧本创作、智能分镜、角色与场景资产管理、图像生成和视频生成。融合无限画布、AI Agent 与可视化工作流，支持本地部署、自选模型、跨平台客户端、MCP 与插件扩展。Open-source AI short drama and motion comic creation tool with an infinite canvas, AI agents, storyboarding, image and video generation. Self-hosted and extensible, with a canvas-based approach similar to LibTV and TapNow.

NEAR AI's open-source agent OS that treats agent execution as a privacy-first harness problem: untrusted tools run in WASM sandboxes with capability-based permissions, credentials are injected at the host boundary with leak detection, and prompt-injection filtering plus endpoint allowlisting constrain agent behavior. The combination of local encrypted storage, hybrid-search memory, MCP server supp

Desktop-grade agent for data analysis, slides, docs, and web research.

基于 AI Agent + MCP 工具链 + 渗透 Skill 编排， 配合大语言模型， 自然语言输入 → 自动完成「信息收集 → 漏洞发现 → 漏洞利用 → 报告生成」全流程。

Bridge between QQ (SnowLuma OneBot v11) and DeepSeek Harness agents: social simulation, safe MCP tools, slang learning and more.

Self-hosted AI penetration testing agent built on LangGraph that maps the attack surface into a graph, exploits it from a Kali sandbox with human approval gates, and opens pull requests with fixes

PentestAgent is an AI agent framework for black-box security testing, supporting bug bounty, red-team, and penetration testing workflows.

Rust - Personal AI gateway with single-binary deployment, multi-provider LLM support, long-term memory, sandboxed execution, voice, MCP tools, and multi-channel

crewAI Tools is a library that provides a framework for developing sophisticated tools to enhance crewAI agents, with methods for subclassing BaseTool, utilizing the tool decorator, and guidelines for contributing to the ecosystem github

Open-source personal AI operating system on Spring Boot + Spring AI Alibaba — one JAR ships a web admin, desktop app (bundled JRE 21), embeddable widget, and 8 IM channels sharing the same agent

Open-source, local-first desktop AI research workbench for scientific computing with Python/R, MCP bioinformatics tools, SSH/WSL/GPU runtimes, and OpenAI/Anthropic models.

A Claude Code skill by Hao (駱君昊) that learns your Facebook voice and auto-posts to FB / IG / Threads / X with a 14-day content calendar. Mega-viral validated: 80K reach / 448 likes / 500 comments on first post. Includes Day 2 flop postmortem.

Lynkr is a proxy that lets Claude Code CLI talk to non-Anthropic LLMs, manage local tools, and compose Model Context Protocol (MCP) servers with prompt caching, repo intelligence, and Git-aware automation and several other features similar to anthropic backend

Send bots to browser-based meetings (Zoom, Teams, Google Meet) to gather live transcripts, speak text and post in the meeting chat

Dulus Ai — Agentic AI, 100% cache hit rate, Dulus is made to save you money [Gui, Web, Cli, Telegram, 2,000 MCP, 100K Skills . LiteLLM (100+ providers), local models via Ollama, /lang in 34 languages, Mesa Redonda, I create the first utility coin that can be used 100% as AI quota or Fuel, is called $Dulus

Turn any video into a Mixamo-rig animation - GVHMR estimator, spec-driven retarget, FK apply in Blender via MCP. Works with any Mixamo character; built to be operated end-to-end by an AI agent.

SSH, SFTP, port forwarding, VNC/RDP and an AI assistant、MCP in one native window — a Go desktop app built on Viva, with no webview.

AI-powered modular Active Directory red-team framework for authorized penetration testing, AD enumeration, attack-path analysis, Kerberos/ADCS workflows, reporting, operator automation, and MCP server integration.

Stop babysitting local AI agents. Just notifications, approve, and resume your Codex,Pi,Grok, or Claude code sessions anywhere. 0-Intrusion mobile control bridge via Telegram/微信/飞书. No hooks, no skills, no MCP.

Open standard for software intent in the AI agent era.

Telegram assistant that turns your messages, voice notes and photos into an Obsidian-compatible markdown vault it remembers across sessions. Crons, skills, MCP and Google Workspace from an in-chat menu. Self-hosted in one command, MIT.

Hanzi is an open-source MCP server + Chrome extension that gives AI agents your real signed-in browser. Task-level automation — one tool call delegates an entire workflow. Built-in skills for LinkedIn prospecting, E2E testing, and social posting. Works with Claude Code, Cursor, Codex, and Windsurf. Setup: npx hanzi-in-chrome setup github

Self-hosted personal AI agent with a built-in WebUI, persistent memory, MCP tools, sub-agent delegation, scheduled automations, and support for local or cloud models

Local-first CLI (MIT) that keeps one reviewed source of the MCP servers, tools, skills, and memory your AI coding agents share, synced into each tool's native config (Claude Code, Codex, Cursor, VS Code, OpenCode, and more) with a review gate before every write and a rollback receipt. No daemon, no lock-in

Self-hostable personal AI assistant built with LangGraph, combining email and calendar tools, persistent memory, MCP integration, and human approval flows for actions

Toolkit that gives Claude Code, Cursor, and Codex structured, step-by-step workflows and an MCP server for building, exploring, and deploying data pipelines with dlt

Run LLM agents over pandas DataFrames

Applies the autoresearch keep-or-revert loop to TPU model performance (MFU / tokens-per-sec) on v6e hardware: profiles each run through an XProf MCP server, makes one model-code change per experiment, and keeps or reverts against measured MFU. Pairs the loop with a Karpathy-style LLM wiki for domain knowledge and per-experiment optimization traces; includes Llama3-8B and Qwen3-8B case studies acro

Crystal - AI assistant with kernel-enforced sandboxing, multi-provider LLM, voice, vision, cron, and MCP tools

Open-source AI study companion powered by Gemini AI. Assignments tracker, SRS flashcards, GPA predictor, AI quiz generator, Pomodoro focus timer, and Claude Code MCP server. Live demo

Native macOS agent that runs agent swarms in parallel, opens terminal sessions it can undo, and reaches your mail, calendar and meetings; runs on a local Ollama model or your own Anthropic key, and any coding agent can drive it over MCP or its CLI.

Package manager for AI agent assets (skills/CLIs/MCPs/loops). Install, lock, and sync skills across Codex, Claude Code, Cursor, Kiro with SAT-based dependency resolution and SHA-256 lockfile. 23 curated enterprise skills included. Available on PyPI

Open-source local memory and task queue for AI coding agents (Claude Code, Codex, opencode, Kimi): decisions survive context compaction, memory is anchored to code and follows it, everything stays in one SQLite file with no API key

Curated x402 directory by 24K Labs. 300+ handpicked entries across facilitators, SDKs, MCP servers, APIs, and tools, with editorial writeups and verified badges for production-confirmed services. Backed by a 29,000+ entry full catalog sourced from CDP Bazaar and Agentic.market

Governance/control plane for AI coding agents. Validates risky actions, records findings, gates approvals, tracks budgets/providers, and creates proof bundles across Claude Code, Codex CLI, OpenCode, Cursor, and other hosts

Mortgage refinance plugin by LendTrain — real-time institutional pricing, compliance, and FHA/VA loan detection via MCP. No API key required

On-chain reputation scoring API for AI agent wallets on Base L2. Returns 0-100 trust scores across 5 behavioral dimensions from transaction patterns and multi-chain attestations. Features x402 micropayments, TypeScript SDK, and ERC-8004 on-chain reputation publication github | docs | npm