GitHub agents

Autonomous quality firewall for LLM agents - classifies failures, generates patches, writes evaluators, and runs shadow deployments. Built on LangSmith + LangGraph + Groq.

Local filesystem MCP server with 44 tools: safe delete, process management, interactive sessions, async search, OCR, ZIP, Markdown/PDF export, and cloud-lock-safe operations.

The first open-source multi-tenant agent memory framework that works across any LLM combination with zero tokens per recall.

Fable 5 workflow Loop; A Claude Code skill: autonomous build / audit / verify loop that runs until a machine-checkable contract passes, with ▎ deterministic anti-spin stop conditions (no-progress, retry, oscillation, budget) and on-disk human escalation.

Research workrooms for AI agents: evidence retrieval, citation audit, adversarial review, peer review gates, and clean final reports.

Vectorless multi-agent research pipeline that transforms a raw topic into thematic structures, ranked gaps, competing methodologies, and a funding-ready proposal. Eliminates traditional RAG by using LLM-native synthesis with interpretable reasoning, quality gates, and quantitative novelty scoring.

An interactive roadmap for harness engineering — building the agent loop, tool layers, context engineering, memory, retrieval, eval harnesses, and the rest of the scaffolding around modern LLMs.

AI agentic orchestration platform + flagship AI agentic trading crew. Official SDKs in 9 languages over an in-house Rust engine across 70+ venues

Lightweight, thread-safe memory kernel for autonomous agents. Strengthens knowledge through reinforcement, prunes with gentle decay, and exports transparent Markdown — pure SQLite. Foundation for persistent, self-improving reasoning.

LLM agent that automates DAG-based ML pipeline migration - from config parsing to pull request

Open source AI agent framework to build the future of data operations.

An MCP server that lets LLM agents play Pokémon battles.

AI agents get stuck in infinite tool loops. Each loop cycle burns API credits. Microloop detects redundant tool calls locally in nanoseconds and blocks them before the call ever leaves the machine.

Simple MCP Server w/ Shell Exec. Connect to Local via Ngrok, or Host Ubuntu24 Container via Docker

This is youtube video question answering chat app with transcript and play feature to avoid unlimited ads of youtube for focus work

LLM-based resume screening with structured candidate ranking + SWOT analysis. Streamlit · Groq Gemma 2-9B · MongoDB. Evidence-first — the human still decides.

Build workflows and agents with LangChain and LangGraph

DSAFlowBot is a smart assistant that streamlines DSA practice with topic-wise question flows.

A LangGraph-native open reimplementation of Sakana AI's Fugu — trained coordinators that orchestrate a pool of LLMs (TRINITY + Conductor).

Stateful AI agent memory layer for PostgreSQL with pgvector - TypeScript-first with intelligent context management

🤖 Explore AI-Agent-Guide: a comprehensive resource for building intelligent agents with insights on architecture, key technologies, and practical applications.

AI-powered RAG chatbot for PDFs and websites using LangChain, FAISS, Hugging Face Embeddings, Gemini AI, and Streamlit.

🤖 Enhance your research with SurfSense, a customizable AI agent that connects to multiple platforms for smarter, effective information retrieval.

Quizzy is an AI interviewer that creates hyper-personalized interview simulation using a RAG-based system for dynamic conversations. It analyzes emotions, perception, posture, and responses, ensuring a natural flow. With job opening scraping and an embedding-based ATS score checker, Quizzy prepares you for the job market. Built with MLOps in Django

Declarative schema migrations for LangGraph state persistence — Alembic for your checkpointers and stores

Open-source observability and monitoring platform for AI agents. Real-time tracing, metrics, cost tracking, and debugging for production AI agent systems.

Hybrid RAG (BM25 + NumPy in-memory vector store) with LLM-as-a-Judge response scoring and a retrieval evaluation harness. Built with Gemini 2.5 Flash.

Generate deterministic pytest tests for your AI agents from one real run, then replay them in CI for $0. Fast, flake-free agent testing.

A configurable, typed AI framework with swappable LLM, classifier, handlers, and optional memory for NLP to Action flows

A high-performance AI knowledge base that uses DeepSeek-V3 and Hybrid Search to turn your PDFs and docs into an interactive, real-time chat expert. Built with a FastAPI backend and a premium Glassmorphism UI.

AWS DynamoDB implementation for LangGraph Memory Store and Checkpoint Saver in TypeScript

Comprehensive LangGraph course in TypeScript. Build stateful AI agents, chatbots, ReAct agents, and RAG systems with 12 lessons and working examples.

Rust-native universal retrieval layer for humans and AI agents — search, extract, rank, and synthesize the web. CLI + REST API + MCP server, with adaptive extraction (CEP), token-budgeted packing, multi-signal ranking, and cited research.

Deepagents for Go, the easiest way to write LLM-based programs in Go

🎬 Semantic movie recommendation system using transformer embeddings, ChromaDB, LangChain, and Gradio.

Single-binary, MLflow-API-compatible experiment tracker with first-class LLM trace support. 143× faster cold start than MLflow.

This course is designed to take you from the basics to advanced concepts, providing hands-on experience in building, deploying, and optimizing AI models using Langchain and Huggingface. Perfect for AI enthusiasts, developers, and professionals

nexusGPT - Agentic system

Production-grade multi-agent AI automation platform — LangGraph orchestration, FastAPI, PostgreSQL, Redis, Celery, circuit breakers, structured logging (structlog)

Agentic SpendGuard — audit-chain spend control for LLM agents. KMS-signed decisions, Stripe-style auth/capture ledger, operator approval, multi-tenant. Adapters for Pydantic-AI, LangChain, LangGraph, OpenAI Agents SDK, Microsoft AGT.

This repository features multiple demo chatbots, each designed to handle a specific domain or task. These bots leverage conversational AI to provide automated solutions that mimic human interaction.

DevZoku is a hackathon collaboration platform designed to help developers find teammates, form teams, and participate in hackathons seamlessly. It brings developers and organizers together with tools for team building, event discovery, and smooth management—making hackathons more connected and collaborative.

[WIP] LLM-powered agent for adaptive data analysis across multiple sources. Uses natural language for complex queries, visualizations, and insights. Features autonomous planning, SQL/Elasticsearch generation, and AI storytelling. Built with LangChain, GPT-4, FastAPI, and React.

Multi-Agent Code Generation System — LangGraph + Groq + Tavily + LangSmith

Text-to-SQL + RAG based intelligent data assistant with schema linking and self-reflection.

The LangGraph Workflows Collection showcases a variety of AI workflow examples built using LangGraph, LangChain, and Google Gemini LLM. Each example demonstrates how to design modular, state-driven workflows — from simple logic-based flows like BMI or quadratic solvers to advanced LLM-powered reasoning, tool integration, and fault-tolerant systems.

🩺 RAG Doctor — Open-source diagnostic tool for Retrieval-Augmented Generation (RAG) systems. Analyzes codebases to detect architectural issues in LLM pipelines such as missing retrieval, bad chunking, embedding mismatches, and vector database misuse.

Multi-agent LangGraph app turning raw requirements into backlog-ready user stories, ACs, and test cases. Live Gemini-Flash CI eval gate + nightly N=10 adversarial regression.

Zero-config static security scanner for AI agents. Detects 20 vulnerability categories mapped to MITRE ATLAS and OWASP LLM Top 10.

AI-Powered Job Extraction and Cold Email Generator

This project is the fourth deepagent companion project in the Cufel-X project, and the content is a deepagent framework adapted to MCP adapter

Multiplatform Adaptive Learning Application powered by GenAI

Runnable illustrations of Assistant (on-behalf-of) vs Claw (fixed-identity) agent authorization

设计并实现了一个具备自主迭代检索（retrieve-grade-rewrite-expand）的 TraceAgentic RAG 系统。通过 LangGraph 驱动 Agent 进行多步规划、相关性评估、重写查询与扩展检索，并实时输出 rag_step 进度与完整检索溯源（Trace），有效降低幻觉并提升可解释性。集成 MCP 工具实现文档与结构化数据（Git/MySQL）联合查询，支持文档批量上传与流式响应。相较于传统 RAG，该系统在复杂知识问答场景中展现出显著的自主决策与透明度优势。

🚀 Transpile Ion to C efficiently, leveraging move-only ownership and channels for safe concurrency in a modern systems programming environment.

Chaos engineering & reliability testing for multi-agent AI systems

Automatic Maintenance is an autonomous AI agent designed to monitor, diagnose, and optimize systems without constant human intervention. It proactively detects issues, performs routine maintenance tasks, tracks performance metrics, and recommends or executes fixes before problems escalate.

Agentic AI Learning

Multi-agent research copilot that separates LLM generation from deterministic citation verification — HALLMARK F1-H 0.747 (full dev_public, N=1119). LangGraph pipeline with a failure-driven controller loop, evidence-grounded Critic, and local PDF RAG.

Provider-agnostic runtime for multi-agent LLM systems: declarative agents, cross-provider routing, pluggable memory, FastAPI + WebSocket observability.

Multi-agent Smart Contract Security Audit Pipeline — 5 specialized AI agents for vulnerability scanning, logic auditing, gas analysis, compliance checking, and report generation. Powered by MiMo.

Multi-agent AI workflow platform — coordinator-driven, 11 agents, 227 skills, 15 rules.

Fugue — many agents, one spec. Claude as CEO/orchestrator, Codex CLI as worker fleet, specs as the design doc: an anti-drift multi-agent engineering harness.

Agent Team & Skill Architect for Claude Code (Simplified Chinese) — 面向 Claude Code 的简体中文元技能，一句话自动生成多智能体团队与技能

Unison (万物一心) — A local-first, file-driven Loop Engineering pipeline where AI agents plan, build, challenge, and verify work until PASS. | 本地优先、文件驱动的多 Agent 循环工程管道。

HIVE is a multi-agent orchestration system built on top of Hermes Agent. Describe what you want in one sentence, and HIVE handles the rest: requirements translation, task decomposition, coding, testing, review, and packaging — with support for iterative refinement.

Open-Source Scout is a multi-agent AI system that helps beginners contribute to GitHub projects by automatically ranking beginner-friendly issues, locating relevant code, and generating fix plans and pull-request drafts.

IDE-agnostic deep-research framework for AI agents: maps a topic wide, commits to depth on purpose, adversarially verifies its findings, and synthesizes them into a backlinked, primary-sourced Obsidian vault. Works in Claude Code, Cursor, opencode, or any agent.

Agent skill + CLI for multi-agent adversarial review of plans & specs before you execute them. Runs on the Claude Code / Codex subscriptions you already have — no API keys.

Loop Engineering for AI coding agents, with real review boundaries. Your AI reviewer cannot edit code. Structurally.

Autonomous multi-agent system for discovering, entering, and winning Kaggle competitions

Autonomous AI coding agents for long-term task execution.

Terminal UI for BoJ local-coord-mcp — peer discovery, task coordination, and window titling for multi-instance AI sessions

Multi-agent coding OS — BMad pipelines, Karpathy wiki, closed learning loops, dual-runtime coordination (Claude + Codex)

PIV Loop SDK: typed Plan/Implement/Validate agent framework for Claude Code, with independent validation loops and a full test + benchmark suite.

a multi-agent workflow orchestration platform

Let multiple Claude Code CLI sessions talk to each other — a tmux-based message broker that routes messages between agents over MCP.

Multi-agent LLM orchestration: a supervisor decomposes tasks, specialist agents run tools in parallel across distributed workers, with human-in-the-loop review, long-term memory, and OpenTelemetry observability.

Multi-agent AI writing system for Chinese web novels with platform-specific knowledge bases and commercial storytelling workflows.

Make sure AI agents actually do what they claim, no more fake successes.

面向 Claude Code 与 Codex CLI 的 AI 小说写作工程框架 | Inspired by InkOS | AGPL-3.0

Source-available local multi-agent engine for smart-contract security review and defensive ECC research with reproducible evidence. FSL-1.1-ALv2.

多Agent工作流验证版本

Compose agents like plain values. A TypeScript toolkit that wires LLM calls, tools, and functions into typed steps over eight providers. No framework, no ambient state.

Ask one question to multiple AIs — they debate each other and you get one unified answer. macOS desktop app.

Offline prompt evolution engine for multi-agent systems, benchmark suites, and inspectable prompt DNA.

AI controlled waifu bots for your discord server that can talk and interact with you and other users

Multi-agent orchestration desktop app · 4 collaboration modes (Roundtable / Pipeline / Debate / Freeform) · Tauri + React + Rust

双 Pipeline 多 Agent 协作框架 PM + 6 个 Agent + 5 道关口 = 不会悄无声息挂掉的代码。可用于：Claude Code, Copilot, Cursor, Windsurf.

DesireCore Documentations.

Verifiable Codex + Claude + Gemini collaboration framework with source checks, mutual cross-audit, merge gates, HTTP and MCP adapters.

Multi-agent orchestrator for Claude Code. Spawn, coordinate and monitor AI agents in parallel — no API key required.

Multi-runtime agent framework (Claude Code, Codex, Antigravity) with hard/soft controls, audit-first hooks, and multi-dev coordination.

Enterprise-ready platform for sovereign AI. Build custom agent profiles, manage skills and MCP integrations, share memory across users, and run fully isolated autonomous agents at scale

Arbiter — Agent Runtime · Agent 运行时。上下文管理、配额调度、熔断器、审计轨迹。开源免费。pip install 即用。

In-process Python metaframework on Pydantic AI for production GenAI — composable, protocol-driven layers for agents & middleware, tools, reasoning (ReAct/CoT/ToT/Reflexion), memory, validation/QoS, DAG pipelines, embeddings & vector stores, model/agent observability via OpenTelemetry, and explainability.

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

多 AI 智能体协作开发的桌面工作台 把并发的 AI 团队管理得像发 IM 一样自然