Python agents

🛡️ The CrowdStrike for AI Agents — Runtime protection, threat detection & security monitoring for LLM agents. Supports LangChain, CrewAI, AutoGen, OpenAI. Python • JS • Rust • Go • Ruby • .NET

ThreatScope analyzes codebases the way a pentester would identifying vulnerabilities, logic flaws, and attack paths before attackers find them.

AI agent governance layer — sign, monitor and control every agent action. EU AI Act · ANSSI · NIST ready.

Examples of using AutoGen agents and agentic frameworks. Requires Docker container.

Your personal AI blogger.

An intelligent interview simulation system powered by AutoGen and OpenAI GPT models. This project creates an interactive interview experience with multiple AI agents playing different roles in the interview process.

Magentic-one AGENTIC workflow system for InsurTech @Insur.Cap

Investigating the usefulness of "AutoGen" by Microsoft. AutoGen is a framework for simplifying the orchestration, optimization, and automation of LLM workflows. This repository contains experiments testing multi-agent coordination, task delegation, and performance benchmarks.

AI Agent for Data Analysis

AutoGen Agents Creation and Testing - VibeCoded by a real developer to enable more people to do stuff using vibes.

A chess face-off between AI Agents using AutoGen. Visualize every move and watch AI vs AI, turn by turn.

AutoGen Magentic-One Sample with security best practices.

AutoGen Practice

Lightweight implementation of the OpenAI open API on top of local models

A modular AI application for creating high-engagement blog posts. This framework leverages a multi-agent system from Microsoft AutoGen to autonomously write, edit, and optimize content, supporting multiple LLMs, optional PDF context, and customizable article length.

cohere-lablab hackathon.

A zero-dependency Python decorator that hides internal arguments from LLM tool-calling schemas via __signature__ mutation.

Rumi GPT is a chatbot built using LLM that provides Rumi quotes to help motivate and inspire you 😇

To examine, interpret and analyze the details and evidence you want from crime files

Unleash potential utilities of LLM models with Llama Index

A FastAPI-based application that combines a Retrieval-Augmented Generation (RAG) system with an intelligent agent for enhanced information retrieval and query handling.

A simple API chatbot that uses LlamaIndex and LlamaParse to read custom PDF data.

A supercomputer simulated by a large language model provides you with a comfortable experimental experience. NOTE: DON'T PUT TOO MUCH FAITH IN NUMERICAL RESULTS.

Create LLM-powered webapps with ease

Query on Videos from YT by providing the URL Drop a YouTube URL, and I’ll handle the rest: - QA on the video - Timestamps & frames for answers - Precise answers delivered

There is a new, more advanced solution in here: https://github.com/hanit-com/azure-openai-conversational-chatbot.git.

This repository is dedicated to the daily research and practical application of AI technologies, especially those revolving around Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and Multi-Agent Systems.

In the project, I will build a mental health consultation system for users based on the DSM-5 data set, using the Llamaindex framework. After consulting the user, the system will compile statistics through the conversation.

Advanced RAG chat with PDF

A simple FASTAPI chatbot that uses LlamaIndex and LlamaParse to read custom PDF data.

An open-source tool using Neo4j-backed knowledge graphs for chatting with your documents

A RAG demo using LlamaIndex that allows you to chat with your uploaded PDF documents

Modular multi-agent AI framework - autonomous agents with memory, tools, and policy management built in Python

🧠 DataForge — An intelligent LLM fine-tuning dataset generator that transforms raw prompts into structured ShareGPT-format training data through a multi-stage AI pipeline. Built with FastAPI + React/TypeScript, featuring real-time progress tracking via SSE, and FAISS-powered deduplication. Just prompt it, and watch your dataset build itself. ⚡

The serialization engine for LLM agentic workflows. A drop-in Python/Rust replacement for JSON that mathematically reduces context window bloat by 44% and deserializes 3x faster. Built for multi-agent swarms, RAG memory, and AI orchestration.

Multi-Agent Grid Environment (MAGE)

Risk-gated autonomous trading with Darwinian strategy evolution, HMAC-signed execution, forward Monte Carlo simulation, reflexivity detection, and multi-exchange orchestration across 5 exchanges. 1,374 tests, security-hardened, production-ready.

A Jupyter Notebook applet to teach and explore Multi Agent Reinforcement Learning - the Nash Q Learning. "Multi agent systems" project @ PoliMi

A compact, polyglot MCP add-on for Claude Code that lets Claude confer with GPT, Grok, Gemini, Mistral, Groq, and DeepSeek — to reason, debate, plan, and peer-review. One config and bounded limits on rounds, tokens, and time.

Cooperation and Exploitation in LLM Policy Synthesis for Sequential Social Dilemmas

Your personal AI assistant - self-hosted, containerized, and intentionally opinionated.

少数派博弈的一个简明Python3实现

Launch Cisco NSO beyond traditional automation. Multi-agent intelligence via AGNTCY + visual workflows via n8n = faster deployments, zero drift, & enterprise-scale network operations.

Autonomous scientific discovery CLI powered by local LLM (Ollama). Generates hypotheses, critiques them, searches the web, runs sandboxed experiments, and iterates - using the ADM-3 multi-agent framework. Fully offline. No API keys.

A multi-agent statistical discriminative sub-trajectory mining technique for analyzing the trajectories of multiple agents using SportVU player tracking data from the 2015/16 season.

Welcome! This guide will walk you through running your Swarms Agent in a Trusted Execution Environment (TEE) using Phala Cloud. This setup ensures your agent runs in a secure, isolated environment.

Declarative, JSON-native AI agent framework

End-to-end encrypted communication for AI agents. The security layer that Google A2A forgot.

Self-hosted deep research engine — multi-agent AI pipeline with 6 LLM providers (incl. local Ollama), 5 search backends, real-time streaming, and a visual pipeline editor. Zero API keys required to start

VibePoster: Editable Posters, Powered by Multi-Agent AI and Knowledge Augmentation

A wrapper for MuJoCo XML files, which turns them into a (Multi-Agent) Reinforcement Learning environment

PM + Dev + QA agents engineered for structure, security, and short context windows.

4 AI agents sharing a Discord server — autonomous coordination, shared memory, and emergent behavior. Built on OpenClaw.

Multi-Objective Multi-Agent RL with non-linear utility functions

Personal AI assistant with 50+ integrations. Email, calendar, tasks, messaging, and more — all working together.

Neural swarming controller models for multi-agent and single-entity simulations.

Intelligent NL2SQL Multi Agent Orchestration framework on Gradio

Autonomous coding agent with swarm orchestration—multiple agents in parallel worktrees. Six modes: greenfield, feature, refactor, fix, evolve, security. CLI + Web UI, approval gates.

Multi-agent code review CLI using NVIDIA Nemotron 3 Super

A real company operated by AI agents, governed by Y*gov

Multilingual incident intelligence — seven AI agents, four languages, two modalities, one shared case state

AI-powered multi-agent system that transforms Telegram into an intelligent automation hub — routing user intent across vision, browser, desktop, and code agents using dynamic model orchestration.

Spin up a Claude-powered multi-agent team from one sentence

Claude Code自然言語マクロプログラミングガイド

🧠 Universal long-term memory for AI agents. GraphRAG-powered knowledge base with vector search + graph traversal. Privacy-first, local-only, MCP-compatible. Connect Claude, Copilot, or any AI assistant.

Managing Coding Agents for Coding Agent

A swarm simulator that uses client-server architecture with socket communication to implement swarm algorithms

A role-card and flow-template driven multi-agent discussion platform for research brainstorming, academic debates, real-time meetings, and structured report generation.

Multi-agent simulation of a public goods game with trust dynamics and reinforcement learning

Local AI agent runtime and CLI with memory, MCP, browser control, automations, and subagents.

Multi-agent collaborative teaching platform with tree-structured conversations

Don't let years of training data sit unused in Strava. An AI coach reads it all, pinpoints why you've plateaued, and builds a safety-gated marathon plan. (Claude Code plugin)

Curtis Bot — Claude Code + OpenClaw bridge for Linux and macOS. Use your existing Claude Code CLI subscription (Max/Pro) as a full AI agent you can text from Telegram. No extra API keys. Phone-to-code, cron monitoring, 50+ skills, multi-agent orchestration, LocalSend file transfers, Git from your pocket.

A simple and powerful package that leverages Streamlit and LangGraph to build interactive, multi-agent web interfaces

⚙️ Dr. Document – a multi-agent 🚀 AI documentation generator using LongCat instances with iterative workflows, specialized agents for code analysis, requirements extraction, and management oversight, plus GitHub integration for automated README generation and validation cycles.

Multi-user MCP gateway for Slack, Teamwork, and Telegram, enabling secure AI tool access through a single endpoint.

智能对话式智能分析平台，让数据分析变得简单自然

A reference architecture demonstrating local-first LLM orchestration for SOC compliance workflows. Built on ~11GB VRAM edge hardware using Ollama, featuring a multi-agent coordination pattern with file-based handoffs, a two-lane integration model (human-initiated vs automated processing), and a HITL approval gate for resource-sensitive operations.

PHAGE — developed as part of Cybersecurity and AI coursework at Esprit School of Engineering. Real-time threat detection using machine learning and a multi-agent AI system.

Multi-LLM debate engine with a shared expert validator that fact-checks each round

Fly-in is a graph-based drone routing simulator that uses capacity-aware pathfinding and real-time Pygame visualization.

A toy local memory system for multi-agent workflows

Multi-agent strategic deception evaluation framework for LLMs using Secret Hitler as a testbed. Analyzes AI reasoning, trust dynamics, and deceptive behavior patterns.

18 AI agents powered by VEROQ — automated trading with fact-checked intelligence, bias detection, and forward predictions.

Constitutional Swarm Mesh — orchestrator-free multi-agent governance

An AI-powered YouTube creator research agent. Topic research, script generation, SEO optimization, and creator style analysis — powered by LangGraph, multi-model LLM routing, and real-time data tools.

Replay Debugger for AI Agents

Simulación de control de tráfico aéreo con agentes SPADE (Python) en donde se coordinan aterrizajes con agentes de Torre y Aviones.

NexQuant is an autonomous AI agent for quantitative strategy research. Built on a multi-agent framework

🚀 An AI-powered multi-agent system using Google ADK that answers queries about rocket launches, weather, and news. It intelligently orchestrates specialized agents to fetch, process, and summarize information for clear, user-friendly responses.

Asynchronous, collaborative multi-agent framework.

Multi-agent AI system for automated social media content planning and generation

The phase 2 of my AI project, which is adversarial search in Pacman game for reaching the best utility and avoiding ghosts. Minimax with alpha-beta pruning and Expectimax is implemented.

ThinkTankGPT leverages LLMs to simulate expert debates on any topic. Users can input a topic of their choice, and the system dynamically generates relevant experts who autonomously engage in a debate. Additionally, users have the option to join the discussion, adding an interactive element to the experience.

3D version of the classic Blocksworld environment for reinforcement learning

Writer skill for Claude Code: from idea to publish-ready article with multi-role review

One model gives answers. Cyber Colloquium gives you a meeting with multi-LLMs.

A self-improving multi-agent retrieval-augmented generation (RAG) API. A query enters, several specialised agents collaborate, and a final answer is streamed back — with full execution traces, structured logs, automated evaluation, regression-blocked prompt approval, and a human-in-the-loop prompt improvement loop.

iKOMA – Autonomous local AI assistant with secure internet search, continuous plan-execute-reflect loops, real-time dashboards, and safety-first guardrails.

An autonomous agent for mathematical reasoning and proof