🤖 AI与机器学习 AI 工具

Deep Researcher AI is a multi-agent system using CrewAI and Linkup to perform real-time, web-based research. It searches, analyzes, and generates cited summaries. Includes a Streamlit UI and MCP server integration for use in IDEs like Cursor.

# README This is an attempt to implement intelligent dialogue functionality using Spring AI. It explores the application of technologies such as RAG (Retrieval-Augmented Generation), Agent, tool invocation, and MCP (Model Control Protocol) for future learning and reference.

This project creates a LangGraph assistant using tool-calling supported by LLM, integrated with a local WhatsApp Mcp server for context-aware message automation.

We will be building a market research agent in AgentChatBuilder where we will cover how to integrate MCP servers like Zapier and other tools with your agents. We will also learn how to create widgets and connect them with your agent's output. In the end, we will deploy the agent to production, evaluate its responses ChatKit for the frontend.

An intelligent math tutor using RAG for step-by-step explanations.

Agentic AI Agent with n8n, OCR, and MCP Server

An agentic AI assistant, that is helping users with math queries to solve and explain them to the user.

A demo of an LLM Agent with tool access using MCP and Groq.

Agentic, multi-tenant RAG chatbot that extends ChatBot-RAG with MCP tools and hybrid retrieval across SQL (structured facts), Vector/RAG (unstructured semantics), and Graph (relationships).

Get personalized tech stack recommendations and insights for your projects.

A smart multi-agent RAG chatbot for Q&A across 6 document formats.

AgentFlow is a TypeScript server for web content ingestion using AI.

🤖 Tiny Tools MCP Showcase - A 4-node LangGraph agent with MCP server integration, Redis memory, and local LLM support

RAG-Based-ChatBot lets you chat with your PDFs by asking questions and retrieving answers effortlessly.

AIVA is an open-source, privacy-first AI assistant running locally on your machine.

MCP AI Agent Demo for Gmail and GitHub integration using Gradio, CrewAI, and Composio.

True Agent2Agent + MCP multiagent workflow

MCP-powered voice agent with DeepSeek

ResearchPilot is an AI research copilot that orchestrates 6 MCP data sources through a custom Docker MCP Gateway using Cerebras ultra-fast inference (Llama 3.3 70B) to deliver comprehensive research reports in seconds.

The ultimate chat AI agent platform

MCP-Hub is a centralized hub integrating multiple APIs (Notion, Slack, GitHub, Supabase, etc.) via the Model Context Protocol (MCP). Initiated by jbmarcilla, it provides a modular, scalable solution that simplifies tool connectivity, enriches AI applications, and fosters innovation in the ecosystem.

Platform to showcase difference with llm responses with and without context , with and withiut use of MCP

A Simple Model Context Server (MCP) For AI Usecases

这是一个用于体验ClineMCP的测试仓库

A MCP playground

提供跨协议AI智能体通信中间件,支持消息翻译、智能路由、Agent发现和安全认证,实现异构系统无缝协作。

暂无描述

DeepSeek全栈开发助手,提供API集成、模型选型、推理优化、本地部署完整指南,加速国产大模型应用落地

通过分析查询复杂度自动选择最优AI模型,跨多个供应商优化成本与性能,降低推理成本同时保证响应质量。

针对AF2/AF3/Boltz-1等蛋白质结构预测结果进行Gate 1-4阶段审计,自动生成质量评估报告和Go/Hold/Kill决策建议。

全自主24/7运行的Aster DEX永续期货交易代理。完整管理交易生命周期,支持括号单、追踪止损等高级功能,无需人工干预。

基于AI的Uniswap v4自动流动性管理器。支持头寸自动再平衡、费率动态优化、止损止盈限价单、Oracle信号监控，通过Chat界面一键操作。

Agentic Web Learning Tool

基于三层架构(基础层-结构层-执行层)设计和优化智能体系统提示词,提升自主AI系统的可靠性和效能,支持多轮对话和复杂任务编排。

基于 MCP 协议的 AI 工具连接框架

使Openclaw智能体能动态生成子智能体处理大规模复杂任务,支持任务自动分解、并行执行与结果聚合,提升系统处理能力。

实时追踪12+AI服务商(GPT/Claude/Gemini等)的token消耗与成本,提供详细账单报告、余额管理、成本预警功能,帮助企业精细化控制AI支出。

mcp server to access changtianML

基于BytePlus Seedance API的AI视频生成工具。支持文本、图像、参考图生成视频，提供查询和管理功能，适合内容创作和营销场景。

智能Agent选择器,自动根据任务描述在146+专业Agent间切换,支持自动适配与手动指定,确保最优执行效果。

为Gemini和其他AI模型提供实时网络搜索能力,支持新闻、股价、赛事等最新信息查询,增强AI回答准确性和时效性。

基于Google Veo 3.1的AI视频生成工具,支持文本或图像输入,快速生成电影级质量视频内容,适合营销、创意、电商场景。

根据BUG描述自动生成工具名和指令列表,支持用户审核后批量生成JSONL训练数据,加速AI模型训练迭代。

为OpenClaw Agent提供分层记忆系统。L0层4行状态快照实现秒速启动,L1层7天滚动上下文保持连贯性,消除Agent冷启动延迟和上下文丢失问题。

基于硅基流动API的国内文生图工具,支持FLUX.1-schnell免费快速生成和FLUX.1-dev高质量模式,无需VPN即可使用。

通过自然语言描述需求，AI自动生成标准化、可复用的Skill结构，支持单个创建与批量模板生成，降低AI工具开发门槛。

暂无描述

自动搜索arXiv论文并生成摘要、热点统计、趋势分析的完整报告,支持定时监控推送最新研究动态。

为AI提供结构化与向量双轨长期记忆，支持语义搜索，让AI跨会话记住用户偏好、决策与上下文，打造真正个性化的AI体验。

A demonstration repository for exploring the MCP world