Python agents

Agentic research on Japanese equities. Competing analysts produce investment theses from regulatory filings.

Local-first AI inference platform for LLMs — private chat, RAG, defense-in-depth security (Daily/Bulbe, encryption at rest, 2FA), on your own hardware. SvelteKit + FastAPI

Asyncio-native Actor framework for Python agent systems — supervision, middleware, pluggable mailbox

ClawAI is a Lightweight, Action-oriented Personal AI Assistant.

🤖A Voice-First Agentic AI for Government Schemes in Hindi. Built with LangGraph (State Machine), Groq (Llama-3), and EdgeTTS for low-latency, CPU-optimized inference without GPUs.

Official Code for "SciCoQA: Quality Assurance for Scientific Paper--Code Alignment" https://aclanthology.org/2026.acl-long.1795

📊 Analyze 小红书 data to uncover market opportunities and user needs, generating automated reports for business idea validation.

A Python package for tracking and analyzing LLM usage across different models and applications. It is primarily designed as a library for integration into development process of LLM-based agentic workflow tooling, providing robust tracking capabilities.

A free (no API key needed) MCP server providing web search capabilities through SearXNG, optimized for LLM agents.

GenAI AWS Solution Architect Agent

A personal‑AI workspace that puts privacy first.

Meeting AI assistant, records, transribes, auto generated summaries, can also use llm agent to send summaries as mail or push to a notion page.

Project that scape all github repositories, read them, analyse with an AI and return you on a discord webhook project that could interest you.

Personal desktop assistant with real-time chat, agnostic support for multiple LLM providers (local and free-tiers included), a full agentic team and local automation with permissions. You can customize it as you wish.

Deterministic regression testing for AI agents

Lightweight journal/memory plugin for Claude Code with zero ML dependencies. Features slash commands, proactive skills, and auto-capture hooks.

🤖 Интеллектуальный мультимодальный AI-помощник с системной интеграцией

Ask any question to your documents, PDFs, or SQL databases using GenAI for quick and accurate answers.

Deterministic mock server for LLM APIs. Test your AI code without spending tokens.

📋uPlan - AI-powered development planning automation tool that generates structured project documents and to-do lists. Streamlines your planning workflow with template-based questions, Supports multiple AI models and offers customizable templates for development planning.

Let your AI agent use X (Twitter) on your behalf — post, reply, search, DMs, bookmarks, and more.

SwiftBar plugin: live Claude.ai session & weekly usage in your macOS menu bar

claude-node lets Python directly run, supervise, and integrate the real local Claude Code runtime. It preserves native CLI capabilities through stream-json, explicit session lifecycle, and process-level control — not another high-level agent framework, but a thin runtime layer for embedding Claude Code into your own systems.

:star: Collection of awesome LLM apps with AI Agents and RAG using OpenAI, Anthropic, Gemini and opensource models

MindOps is the layer where AI meets awareness, where observability data becomes self-interpreting, and where signals evolve into neural decisions. It is the operating system for understanding — not just monitoring — modern infrastructure.

Branch-based LangChain learning path for building agentic AI apps, ReAct agents, tool calling, RAG with Weaviate, and a Streamlit documentation assistant.

TalentScout is an AI-powered hiring assistant provide intelligent recruiting and candidate evaluation support. The system features configurable AI personas, adjustable conversation memory, and a user-friendly interface designed to streamline the hiring process through natural language interactions.

Adaptive Malayalam AI tutor for neurodivergent learners with personalized learning and mastery tracking.

The missing developer tool for working with vector databases. A comprehensive desktop app for visualizing, querying, and managing vector data.

A knowledge aggregator for computational biology

多模态智能体RAG客服

Enhance your RAG with Contextual Retrieval

A Streamlit-based Retrieval-Augmented Generation (RAG) application that enables searching within PDF files using Ollama and DeepSeek R1. It leverages FAISS for vector search, HuggingFace embeddings, and LangChain for document retrieval and response generation in Arabic.

Turn any Java, Kotlin or Python project into an AI-ready context file in one command.

An AI technical biographer that transforms your GitHub footprint into a professional Persona Audit PDF. Features agentic repository deconstruction and a savage roast mode.

Benchmark-driven model routing for OpenClaw with data-driven policy tuning. Inspired by karpathy/autoresearch.

Turn AI into anyone — data-driven persona generation from real social media. 说一个名字，AI 就能变成 TA。自动采集真实数据，生成比手写人设真实100倍的角色。

Connect your camera to an AI agent for hardware automation. Consent-gated webcam plugin for Claude Code + Codex CLI — point at a breadboard, multimeter, 3D print, or paper; the agent grabs a frame and answers.

🐧 is currently a work in progress... Stay tuned!

Autonomous AI coding agent — local-first, any language, any LLM. 24/7 daemon with task queue, heartbeat monitoring, cron scheduling, multi-agent roles, and web dashboard.

An MCP server that exposes OmniFocus to AI agents via the Model Context Protocol (MCP).

AI-driven Android UI test automation — plain-language cases run by an agent on real devices over any network (no ADB/scripts), with visual reports & self-learning.

Get personalized advice from 5 product management experts by analyzing Lenny's Podcast transcripts

Persistent memory layer for Claude Code — Obsidian vault health, connection discovery, brain dump, session notes, email bridge. Five lightweight skills, CLI-first, any taxonomy.

Flask-based web client for Letta AI agent framework with chat interface and agent management

🧠 EdgeBrain — Lightweight edge AI inference engine. Run ML models locally with sub-100ms latency. No cloud. No API keys. Just fast, private AI.

A Python MCP Server that converts ComfyUI workflows into parameter-configurable tools for AI agents. | 将 ComfyUI 工作流转化为参数可配置的 MCP 工具，供 AI 代理通过自然语言驱动图像与视频生成。

A general-purpose AI assistant

Local-first AI coding and automation agent with automatic Ollama setup, hardware-aware models, RAG, MCP, and a friendly terminal UI.

Thousands of templates to get you started. Simplify your automation tasks with ready-made solutions tailored to your needs. Find workflows for Telegram, Discord, Gmail, AI, and hundreds of other integrations. Build your first workflow,

AI agent memory using Modern Hopfield Networks — no LLM calls, no database, one matrix multiply. MCP server for Cursor, Claude Code, and other AI coding agents.

Memory infrastructure for AI agents!! store, recall by meaning or association, ingest documents, and track beliefs in one Python library.

LiAgent OS is a local-first AI agent OS for building a private personal assistant with governed autonomy across local models and hybrid cloud services. It brings conversation, tool use, multi-agent orchestration, task scheduling, heartbeat execution, human approval, and auditability into one long-lived loop.

Zero-config visual debugging and auto-evaluation for LLM agents. Local-first tracing with a beautiful dashboard for OpenAI, LangChain, CrewAI, and more.

Do MCP tools serialize in Claude Code? Empirical study: readOnlyHint controls parallelism, IPC overhead is ~5ms/call. Reproduces #14353.

Local acceptance gate for coding agents: blocks false-green done until task-scoped proof is rechecked.

Zenodo MCP server — tool-based LLM integration with Zenodo (open-access research repository) via the Model Context Protocol (MCP)

Colombian SME Agent: LangChain+Flask chatbot that scrapes DIAN, MinCIT & CCB pages, chunks and embeds text into ChromaDB, then filters by similarity to deliver structured financial, tax, and growth recommendations.

Production-grade AI multi-agent collaboration framework builder Skill for Manus

A comprehensive collection of production-ready AI agent applications featuring single agents, multi-agent teams, and autonomous game-playing systems built with Phidata, CrewAI, LangChain, and Google ADK using OpenAI, Anthropic, Google, and open-source LLMs.

Hacker Agent 是一个专业的渗透测试 AI 助手，具备 OSCP、CEH 等专业认证知识体系。通过集成 MCP 服务器调用 Kali Linux 工具，实现从信息收集到权限提升的完整渗透测试流程。项目采用 ReAct (推理-行动) 方法论，让 AI 能够进行多步骤推理和决策，系统化地发现安全漏洞。

Production-ready Haystack/LangChain pipelines for Hybrid & Parent-Child Retrieval, Diversity Filtering, MMR, Metadata Filtering, Reranking, Query Enhancement, Multi-Tenancy, Agentic RAG across Pinecone, Weaviate, Milvus, Qdrant, Chroma with dataset loaders, evaluation tooling, and benchmarks for TriviaQA, ARC, PopQA, FactScore and Earnings Calls.

一个专注于健康管理的AI智能对话平台，用户可以通过文字、文件上传等方式与AI助手进行健康咨询。系统支持多会话管理、智能文件分析、健康报告解读、RAG知识库检索增强等专业功能。

Multi-agent pipeline that turns any topic into a narrated Manim animation

Enterprise RAG pipelines with native IRIS vector search. 6 production implementations with RAGAS evaluation, LangChain, AWS/Azure configs. No external VectorDB required.

Spec Crew is a spec-driven CrewAI toolkit for Claude Code and OpenCode, with canonical commands, agents and skills.

Multi-agent research CLI. Plan, search, scrape, synthesize, review -- all from your terminal.

🚀 Build and explore multi-agent AI workflows with ready-to-use projects for document serving, Q/A bots, and orchestration.

A2A v0.3.0 协议的 Python 实现 — 让 Hermes Agent 互通

Autonomous AI desktop agent powered by local Ollama LLM — automates files, Excel, web, shell and GUI tasks on Windows.

Composable Python SDK for building single-agent and multi-agent AI systems.

A comprehensive guide for modernizing mainframe applications (IBM z/OS, Unisys, GCOS, ACOS) using Azure AI Foundry with GitHub and Azure DevOps, preserving existing investments while enabling modern DevOps and AI capabilities.

Rare AI Archive: open-source agentic diagnostic AI for rare genetic diseases — decentralized post-training, clinician validation, federated deployment

Paper-driven molecular-dynamics framework with an AI skill that turns physics papers into runnable Taichi simulations.

Open-source hunt playbooks and a policy-driven Sentinel reference framework for detecting runaway AI agents, unauthorized tool calls, behavioral drift, DNS exfiltration, and poisoned MCP tools. MITRE ATLAS mapped.

A Python SDK for building LLM-powered multi-agent systems using the Blackboard Pattern.

Contract-driven, agentic system for deterministic CAD generation from natural language using headless CAD

GraWiki is an open source Python project for graph-backed knowledge extraction, retrieval, and agent memory. It combines two closely related ideas: GraphRAG-style ingestion and retrieval over extracted entities and relationships. Andrej Karpathy's "LLM Wiki" style memory, where prior agent work is stored as durable graph state instead

Run large Mixture-of-Experts LLMs that exceed system RAM on Apple Silicon by loading only router-selected experts from SSD with MLX. Includes OpenAI/Anthropic-compatible serving for local agentic coding.

Just an AI assistant for termux users with coding assistant abilities

MCP Server for AnyLogic Cloud simulation platform integration with AI assistants

Drop-in OpenAI and Anthropic API server that routes through Claude Code

Your Claude Code & Cowork conversations are sitting on Anthropic's mercy — until they're not. claude-backup exports everything to local Markdown, and rescue bundles it for handoff to ChatGPT, Cursor, or any AI agent when you need a Plan B. Lazy install: paste PROMPT.md into any AI

Financial Market Intelligence MCP Server — stock quotes, technical analysis, crypto data, and portfolio insights for AI agents

Comprehension-driven development: Claude quizzes you on your own PR diff so you catch the gap between what you think your code does and what it actually does — before you merge.

High-fidelity PDF/DOCX/PPTX to Markdown conversion with multi-stage QC, image extraction, and Claude Code integration.

An AI-powered backseat coach to fix your skill issue and/or ruin your day :). Supports popular models from OpenAI, Anthropic and Google and self-hosted. Customizable prompting and voice cloning thanks to ElevenLabs and Coqui TTS.

MCP server for the Instagram Graph API — 24 tools for posting, comments, DMs, and insights.

An AI-powered Road Damage Assessment system using YOLOv11s & SAHI to detect potholes and cracks with high precision. Featuring a Streamlit dashboard for real-time analysis, it provides engineering reports based on IRC standards. Integrated with RAG technology to offer expert maintenance consulting derived from civil engineering manuals.

面经爪 · 大厂面试经验库 + RAG 智能问答 | OpenClaw Agent Skill — 牛客爬虫 + ChromaDB + LLM 备考建议

A Python library for Retrieval-Augmented Generation (RAG) that extracts text from PDFs and performs OCR on images. It uses Llama 3.2 for embeddings and answer generation, FAISS for similarity search, PyMyPDF for PDF processing, and pytesseract for OCR.

Search your own documents locally. No server, no sign-up, no data leaves your machine.

Coding agents forget your repo. mcp-brain is the missing memory layer — repo-aware, team-aware, lifecycle-aware. 63% Hit@10, zero LLM cost. Works with any MCP client.

Build Dify workflows from business requirements with design, YAML generation, review, refactor, validation, nanny-mode guidance, and URL-driven app recognition.

Build Your Own AI Employee. The complete starter kit for OpenClaw + Hermes Agent. 155 deduplicated skills, drag-and-drop workspace, case studies, install scripts.

Open-source territorial intelligence platform — analyze French public data (SIRENE, DVF, INSEE, BODACC) with AI agents. Python/FastAPI + Next.js. Self-hosted.

Multi-agent local LLM orchestrator for LM Studio: planner→coder→reviewer, SQLite queue, memory-safe model swaps. GPL-3.0 · CLI: lao

Automate authorized pentesting with LLMs, combining scanning, RAG, and exploit research for faster target analysis, vuln discovery, and reporting

Hindsight Benchmarks Results