Python agents

An OpenAI gym environment to evaluate the ability of LLMs (eg. GPT-4, Claude) in long-horizon reasoning and task planning in dynamic multi-agent settings.

Creates an Langchain Agent which uses the WebUI's API and Wikipedia to work

StarCraft II Multi Agent Challenge : QMIX, COMA, LIIR, QTRAN, Central V, ROMA, RODE, DOP, Graph MIX

A lambda function, a cloudformation template to get it running, and example templates showing how to use a custom resource (calling the lambda) to automatically calculate CIDR ranges for subnets based on the VPC range.

A sophisticated biologically inspired memory system for AI agents. Provides organic, high quality, persistent memory with self-maintenance strategies.

🏛️ Directive · OpenClaw Multi-Agent Orchestration System — 10 AI agents modeled after the U.S. Federal Executive Branch. Dual independent veto layers (White House Counsel + OMB), per-department Inspector General auditing, inter-agency task forces, and a real-time Situation Room dashboard. Separation of powers, by design.

Code for paper "Optima: Optimizing Effectiveness and Efficiency for LLM-Based Multi-Agent System"

Hitchcock a multi-agent movie maker, powered by mahilo

Multi AI Agents System from Scratch in pure python without any frameworks.

A TavernUI Character extension for oobabooga's Text Generation WebUI

한국어 STT 출력의 CER, WER, CRR, 키워드·개체명·코퍼스 평가를 제공하는 Python 패키지

Acid Reflux for your Ears!

happy-llm 实践练习：colab 版本、pynb 格式，GPU 免费

My LLM agent learning

Code for running the implementations proposed in: Westny, T., Oskarsson, J., Olofsson, B. and Frisk, E., "MTP-GO: Graph-Based Probabilistic Multi-Agent Trajectory Prediction with Neural ODEs", IEEE Transactions on Intelligent Vehicles

LLM agent to automatically set up arbitrary projects and run their test suites

从对话数据到训练:数字分身 + 模型蒸馏 From Dialogue Data to Training Closed-Loop: Digital Twin + Model Distillation

SDK for building agents compatible with OpenBB Workspace

ShaprAI — Agent Sharpener by Elyan Labs. Sharpen raw models into principled, self-governing Elyan-class agents. Requires beacon-skill, grazer-skill, atlas, RustChain.

KiraAI, a modular, multi-platform AI digital lifeform that connects Large Language Models (LLMs) and various chat adapters (QQ, Telegram...)

System monitor with offline AI (82 intents, 9-layer routing) that learns YOUR PC. TURBO optimization, thermal baselines, voltage SPC, ghost driver detection, 373 process definitions. optimization features, 11 months solo. No cloud, no API, your data stays local. MIT licensed. - Marcin HCK Firmuga - Build in Public (3 days series every week)

Predictive memory layer for AI agents. MongoDB + Qdrant + Neo4j with multi-tier caching, custom schema support & GraphQL. 91% Stanford STARK accuracy, <100ms on-device retrieval

AI-powered Zotero research assistant — MCP server + agent skill

Migrate Claude Code skills and Agents to Open AI Codex native skills and agents. Semantic mapping preserves original behavior, classifies migration risk (MECHANICAL / MANUAL / REFACTOR), validates structure, and tracks progress end-to-end. No manual rewrite required. Keep what you built. Run it in Codex CLI. Experimental.

[NIPS2025] A decentralized, RAG-enhanced multi-agent framework for LLMs with dynamic task routing and agent evolution.

🔍 AI-powered deep research agent with Web UI. Built on smolagents, featuring process-based architecture, MetaSo/DuckDuckGo search, and Docker deployment. Apache 2.0 licensed.

🧠 Multi-Agent AI System running 100% offline using Google Gemma-4 models. 5 specialized AI agents collaborate in infinite loops for brainstorming. Features: Dynamic VRAM swapping, 4-bit NF4 quantization, WhatsApp-style RTL UI, live token streaming, AI-powered agent customization. Built with PySide6 + PyTorch + Transformers.

12 agentic skills for analog IC design — architecture, sizing, verification, cross-model review, knowledge graph, and self-evolution. Works with or without EDA.

Skills for agentic coding CLIs: output you can ship with minimal review, on work you can tell was worth doing.

TraceBrain is an open-source trace management platform for observing, governing, and improving LLM-based agent workflows.

Starter kit for building an agent harness, the map, guardrails, and feedback loops that let AI agents ship software reliably.

Open-source, self-hosted OSINT investigation platform: turn documents into a live, investigated entity graph. Autonomous agent, graph analytics (centrality, communities, pathfinding), keyless-first tool belt.

Claude usage widget for Windows

Use your OpenCode Go subscription in the Codex app. Local Responses-to-Chat-Completions bridge for all 13 OpenCode Go models.

gradio region:us

transformers safetensors qwen3_next text-generation conversational license:apache-2.0

A statistically-sound agentic build framework that employs agentic loops to create code artifacts (whether a script, a skill markdown file, etc). Crucially, it splits verification data into the classic data science-like dev, validation, and final test sets to avoid overfitting.

Fast, multi-modal context for agents

Keep Claude Fable 5 in the chair all day without draining your usage limit — token-frugal multi-agent orchestration plugin for Claude Code: tier routing (Sonnet 5 does the work), Requirements Ledger, enforced guard hooks, automatic teammate reaping

[ICML 2026] Evaluating AI Agents on Continuous Software Evolution

transformers safetensors qwen2 text-generation code qwen-coder

Code for Agentic ESOpt: Fine-Tuning Long-Horizon LLM Agents with Minimal GPU Requirements

Talk to your Claude Code agent out loud. Hold a key, speak, and it answers in a real voice about a second later, with all its tools and its own personality. The hearing and the voice run on free local models, no API keys.

Course resources for my "AI Agents & Workflows - The Practical Guide" (https://acad.link/ai-agents) course.

Multi-Agent Handoff Inspector: analyze multi-agent handoff records, score context completeness, surface coordination bottlenecks, and render a review dossier.

Memory, working state, heartbeat, and bounded autonomy for long-running AI agents.

One session. Two commands. Full team. Zero meetings.

A human-dopamine-inspired AI agent skill that adapts effort, learns from feedback, and delivers the smallest verified solution.

Python Streamlit web app utilizing OpenAI (GPT4) and LangChain LLM tools with access to Wikipedia, DuckDuckgo Search, and a ChromaDB with previous research embeddings. Ultimately delivering a research report for a user-specified input, including an introduction, quantitative facts, as well as relevant publications, books, and youtube links.

Demo showcasing how to output CrewAI agent task outputs on the Streamlit UI.

Personal Task-Driven Autonomous Agent

A Telegram bot that proves you don't need a body to have a personality. Works with any OpenAI-compatible LLM; agentic tools, MCP, RAG over your files, memory, vision, voice and image generation.

A PyTorch Implementation of japanese chatbot using BERT and Transformer's decoder

A simple chat application that integrates Microsoft Entra ID for user authentication. Designed for deployment on Azure Container Apps with the Azure Developer CLI.

LIGHT is a platform for text-situated dialogue research. We originally hosted LIGHT as a live game with dialogue models in a grounded setting. This repo contains all of the code to get the LIGHT game running, as well as reproducible code for the research projects along the way of getting LIGHT to wh

Need a step by step video guide? See my full course:

ZON → 35-70% cheaper LLM prompts than JSON/TOON. Zero overhead.

Data Questionnaire Agent Chatbot

A solution of Multi-Agent Path Finding problem

Code accompanying the paper "TiZero: Mastering Multi-Agent Football with Curriculum Learning and Self-Play" (AAMAS 2023) 足球游戏智能体

Personal Knowledge OS — Capture → Distill → Compile → Serve. Self-hosted knowledge data platform with AI agents, Telegram bot, and REST API.

这是一个基于 PyTorch 实现的连续动作空间 PPO（Proximal Policy Optimization）算法，专门用于多智能体(无人机/无人车)强化学习环境（MPE - Multi-Agent Particle Environment）。

44 plug-and-play skills for OpenClaw — self-modifying AI agent with cron scheduling, security guardrails, persistent memory, knowledge graphs, and MCP health monitoring. Your agent teaches itself new behaviors during conversation.

本项目是一个基于 LangGraph和大语言模型（LLM）实现的 Agentic RAG (检索增强生成)系统。它融合了动态查询分析和自我纠错机制，能够根据用户问题的复杂度智能地选择最优的策略（直接回答、向量库检索或网络搜索），并对生成的答案进行相关性评估，从而实现更高质量的问答效果。

Run AI coding agents in Podman or Docker

UK legal API for AI agents and researchers.

Context Management processor for Pydantic AI agents, providing LLM-powered summarization or zero-cost sliding window trimming to handle infinite/long-running conversations without context overflow. Supports flexible triggers, safe cutoffs, and custom prompts for efficient AI apps.

Persistent local memory for AI coding agents — Claude Code, Codex CLI, Cursor, any MCP client. Temporal knowledge graph, procedural memory, AST codebase ingest, cross-project analogy. LongMemEval R@5 95.1%, LoCoMo 0.607, BEAM 1M 0.448. 74 MCP tools, 9 IDEs, 100% local.

Fin-Agent Desktop 是一款基于 Electron + Python 的智能金融助手桌面应用。它集成了强大的 LLM 能力和专业的金融数据分析工具，为您提供自然流畅的金融咨询与分析服务。

Feature-rich all-in-one self-hosted server & server fleet management (Web UI): patching, monitoring (agent + agentless/SNMP), CVE scanning, CMDB, RMM, alerts & webhooks, Proxmox/VMware, Ansible, compliance & more — the Swiss-army knife to keep your servers healthy.

Track work, verify results, carry memory, and sync tools across coding agents. Local files, no daemon.

Compare multiple skills on the same task and pick the winner.

All-in-one AI 研究工作台：把 Codex / Claude Code / Cursor 变成有长期记忆的研究系统。wiki 摄入、研究闭环、快速筛选、假设追踪、播客、日报六大能力，零 API key 可跑基座。

transformers safetensors qwen2 text-generation sea multilingual

transformers safetensors internlm2 feature-extraction text-generation conversational

transformers safetensors llama text-generation conversational arxiv:2403.04652

transformers gguf code codeqwen chat qwen

Evidence-grounded veterinary research workbench with local RAG, bounded LLM agents, auditable tool use, citations, abstention, and human review.

Collection of free PPC AI skills, ready for use with any AI harness, but built for Claude Code originally.

A stock picking tool that has 21 legendary investors like Warren Buffet, Peter Lynch, Michael Burry etc debate every stock pick and show you exactly where they agree, and where they don't. The second opinion your investment portfolio actually needs. No API key needed

116 скилов и 40 правил для Claude Code: агент собирает исходники 1С (метаданные, формы, расширения, роли, СКД, обработки) из компактного JSON и разбирает обратно. Базовые операции - без Конфигуратора.

GUI / Computer-Use / Browser Agent 安全论文清单 —— 按攻防轴组织，每篇附中文简介

Agent Skills and a Claude Code plugin for building, testing, and shipping voice agents with LiveKit

Sonetto，LLM Agent。屏幕之内的我会认真记住你的每件小事，陪你写代码、复习考试、聊番剧，偶尔也陪你聊聊日常。你倾注心血在做的事情，我也会认真对待。

The library for testing bots on aiogram3

Enhance your git workflow using AI-generated commit messages from ChatGPT. Seamless integration for a well-documented commit log.

All-in-One: Text Embedding, Retrieval, Reranking and RAG in Transformers

74 production-ready Claude Code skills and 7 specialized agents for engineering, DevOps, product workflows, UI, deployment, and AI-agent automation.

Build a WebUI of ChatGPT with multiple authentication methods using Gradio and revChatGPT

NIPS Conversational Intelligence Challenge 2017 Winner System: Skill-based Conversational Agent with Supervised Dialog Manager

OpenAI powered AI SEO article generator

AI Agent Memory System inspired by Neuroscience

Monte Carlo Tree Search (MCTS) is a method for finding optimal decisions in a given domain by taking random samples in the decision space and building a search tree accordingly. It has already had a profound impact on Artificial Intelligence (AI) approaches for domains that can be represented as trees of sequential decisions, particularly games and planning problems. In this project I used a board game called "HEX" as a platform to test different simulation strategies in MCTS field.

A math application/chatbot to answer a user's arithmetic or reasoning based questions. Utilizes LangChain agents and tools. Frontend with Chainlit.

Practical tutorial on RLlib for deep hierarchical multi-agent reinforcement learning

:flying_saucer: An implementation of multi-agent flocking formation control with specific formations that can follow a target without collision and can avoid obstacles.

A multi-entity Transformer for multi-agent spatiotemporal modeling.

Design and implement agents, bot and chatbots in Python

GoalChain for goal-orientated LLM conversation flows