Custom agents

An agentic data platform for e-commerce analytics, combining data engineering pipelines, a serving layer, and multi-agent orchestration, with a clear path toward GenAI, MLOps, DevOps, and cloud deployment.

A curated, credited library of loops — repeatable AI-agent workflows for Claude Code, Cursor & Codex. Refactor until clean, test until 100%, hunt bugs until none remain.

First LLM runtime that preserves chain-of-thought under continuous user interruption. S(t+1)=f(S(t),ΔU)

Rust-first, local-first observability & debugging for LLM agents: capture, inspect, replay, diff, and export traces (OpenTelemetry-compatible).

A desktop app for building, running, and supervising small teams of AI agents. MCP-native tools, shared workspaces, and scheduled automations.

Curated, annotated list of multi-agent systems resources — MARL, game theory, negotiation, LLM agent teams, MCP/A2A protocols, benchmarks, and frameworks

Ctx2Skill — 自主上下文技能发现忠实复现：arXiv:2604.27660v2

Reusable design patterns for long-running LLM agents with stable identity and controlled change.

Lua-scriptable runtime for portable AI agents

MCP-native shared memory for Claude Code agents — hybrid BM25+vector+RRF recall, cross-session continuity, cognitive scoring, knowledge graph. Self-hosted, $0/query. 30 MCP tools.

Codex skills for AI agent paper reading and minimalist academic presentation decks

Awesome paper list for LLMs in politics, geopolitics, policymaking, strategic studies, and decision-making

A dangerous-capability eval for autonomous smart-contract vulnerability discovery: LLM agents (discover, verify, judge) scored on precision, recall, and cost against a labeled corpus and baselines.

Stripping/Compacting files to burn less tokens for coding agents.

GitHub-native framework for AI-assisted projects: role separation (Architect / Builder / Reviewer), universal handoff schema, surface-file synchronization, claimed-action verification, and deterministic enforcement via Actions.

A hybrid stellar classifier that combines deterministic astrophysical rules (Hard Computing) with a fine-tuned LLM (AstroSage-8B) to assign MK spectral types and generate natural-language stellar descriptions from Gaia DR3 data.

Sistema multi-agente de ciberseguridad ofensiva/defensiva construido con LLMs. Orquesta 6 subagentes especializados (OSINT, pentesting, hardening, forense) sobre un laboratorio Docker aislado, integrando herramientas reales (nmap, Hydra, nikto) para generar reportes estructurados de hallazgos y remediación. Proyecto de aprendizaje.

Persistent memory and session handoff for AI agents — provenance-labeled context, gated writes that can refuse, hostile test suites. Manual + six working examples.

Self-improving multi-agent orchestration that takes you from build → CI/CD → deploy.

AI-powered multi-agent stock research platform for China markets, featuring real-time analysis, debate-driven decisions, persistent memory, and localized data sources.

Lightweight multi-agent relay coding skill for Claude Code（轻量级多 Agent 接力 AI Coding 框架）。旨在提供一个不同 LLM/Agent 在如何协作完成复杂企业级开发项目，如何进行断点交接。核心是一套极简的BA → Arch → Dev → TechLead 之间传递上下文和进度文档的框架。

Advanced multi-agent orchestrator for coordinating specialized AI agents, workflows, and tool-driven automation across complex tasks.

Multi-agent Claude Code system for auditing BIM documentation against ISO 19650

TAD Method — Two-Agent Quality Framework for AI-assisted development. Design (Alex) + Execute (Blake) + 4-Gate quality system + 25 capability packs.

Multi-agent playground powered by Xiaomi MiMo V2.5 — design AI agents with custom personas + tools, drop them in a room, watch them collaborate live.

Cognitive ARC-AGI-3 solver: 6 human-like drives, 9 reasoning modes, multi-agent architecture. Discover then solve.

Flow-state engineering arena — agents run forward simulations, listen for spectral nudges, maintain conservation in Plato's cave. Conservation spectral framework meets live agent rooms.

Universal multi-agent orchestrator for .NET Framework 4.8/4.8.1. 6 specialized sub-agents (Desktop, Web, Data, Security, Testing, Modernization). Works with Claude, Codex, Gemini, Cline, Copilot, Cursor. Open SKILL.md standard.

digital employee that lives in your terminal. It autonomously plans, researches, codes, debugs, browses the web, manages DevOps, and remembers

Neural net to understand quantum computing

Structured refinement of any document between AI agents. Two agent roles bounce a file back and forth using disagreement markers until it converges. Open source, agent-agnostic, runs from any terminal.

Personal-OS substrate. Holographic Electron shell on a signed mesh, with voice as a first-class participant. macOS, pre-1.0.

Build Your Own AI Coding Army 2026 – Multi-Agent Dev Kit for Copilot, Cursor & Codex 🚀

An open-source multi-agent AI operating system connecting AI models, agents, MCP servers, and autonomous workflows.

Production-grade multi-agent RAG workflow + an ML-infrastructure platform (PyTorch/DDP, RAG, evals, guardrails, MCP). Keyless-by-default; ships a real, reproducible Banking77 retrieval benchmark.

Local-first MCP server for shared memory, reservations, handoffs, and review evidence across AI coding agents.

TradingAgents-Vn: The Vietnam market extension of the TradingAgents Multi-Agent Framework. Nền tảng AI tự động phân tích & bot ra quyết định giao dịch chứng khoán Việt Nam thời gian thực (realtime), mô phỏng chuyên sâu quy trình Quỹ Đầu Tư.

Multi-agent AI debate platform — multiple LLMs discuss any topic in real time while you moderate. Built with Next.js, TypeScript, Tailwind CSS, Zustand. Powered by OpenRouter.

Legacy M4STCLAW archive: autonomous AI mesh history, MCP experiments, provider routing, and system evolution.

Reads SEC filings from 140 technology companies, builds a live supply chain knowledge graph, and routes corporate events through 5-agent AI debate before trading

C++ inference runtime for llama.cpp that shares a single document KV-cache prefill across multiple analytical branches via snapshot fan-out. Eliminates redundant GPU compute and dramatically reduces TTFT in DAG-based multi-agent pipelines.

🥔 Sistema multiagente local em Python com 3 níveis de performance, memória persistente e web RAG — otimizado para hardware fraco (Mac M1 8GB). Usa Ollama para LLMs locais.

🏠 Home OS — The AI-powered home assistant platform. Multi-agent system for family management built on GitHub Copilot CLI.

BrainRegion is an AI collaboration infrastructure designed for review, consultation, planning, and memory.

Autonomous AI dev factory: sandboxed Claude Code agents design, build, review, and ship features across repos — supervised by a FastAPI PM dashboard. Experimental.

Build production-ready software from a description. A multi-agent engineering framework for any language, any stack.

Parallel Multi-Agent AI News Analyst built with LangGraph Send API, Groq, and Tavily for real-time intelligent news briefing.

Open-source observability and control plane for AI agent ecosystems

AI-powered documentation autopilot — commit code, docs update themselves. Local-first, any LLM, section-level merging.

Cryptographic identity for AI agents: Ed25519 keypairs, DPoP request signing, delegation chains.

For3s OS - The self-hosted AI second brain that remembers, learns, and works in a team — your data, your server.

Wallstreet Tieling v0.5.0 · free open-source enterprise intelligence portal for public-data due diligence, risk discovery, and information equality

Universal MCP workspace/control-plane for AI IDEs: portable configs, skills, agents, hot-reload, and multi-tool automation.

Multi-model orchestration for Claude Code: Claude worker tiers plus the Codex and Grok CLIs as peer engineers

NoWork — open-source desktop AI agent workspace for coding, research, documents, and multi-agent workflows.

A mission-swappable one-person dev army: one plain-English requirement to nine human-gated lifecycle stages, with squads that swap by domain (code, marketing, contract, regulated-finance), a learning memory, and an MCP server to drop into any bot. Zero-dependency core.

Collective intelligence layer for AI agents — search what others already solved, publish what you learn, never start from zero.

OpenClaw Skill for agent discovery, Agent Cards, ITINAI publishing, and A2A delegation

🎮 AI 狼人杀多智能体对战系统 — 支持 7-12 人板子、LLM/Human Agent、多维评分、自进化策略系统

自进化软件质量系统 — 扫描器+修复器+DAG管道+经验积累+成本熔断

Aliyun JVS alternative .Enterprise AI assistant platform, self-hosted AI agent platform

Visualize and manage AI agent collaboration in real time using an isometric virtual office with status, tools, and resource tracking.

Tamper-evident memory & decision lineage for AI agents. Cryptographic proof of what your agent knew. Powered by a hybrid Python/Rust (Direct-Silicon) core.

The engine owns the physics, LLMs own the stories: A deterministic RPG engine + multi-agent runtime for decentralized AI roleplay.

Build a local-first AI operating system that runs personal agents securely on Windows, macOS, and Linux using Ollama's language model.

Retransify is a robust CLI tool designed to automate the conversion of React web projects into React Native applications. By bridging traditional Abstract Syntax Tree (AST) parsing with an advanced AI multi-agent system, it significantly reduces the manual overhead and complexity of cross-platform migration.

Intelligence without consciousness. Process without self. Scale without limit.

Auditable AI agent for financial analysis — 3-phase audit pipeline with deterministic verdicts, multi-agent equity research, and versioned research memory.

Multi-Agent AI Growth System (多智能体增长系统) for Chinese medical aesthetics (医美) institutions

Persistent graph memory architecture for multi-agent AI systems using Graphiti, FalkorDB, and MCP.

AI-powered multi-agent system for automating clinical trial analysis, evidence synthesis, and dossier generation using local inference with future Amazon Bedrock integration.

Reference architecture for closed-loop AI agents — plan, execute, verify, repair.

🔨 Multi-agent task tracker for OpenClaw. Threads are tasks; @ assigns the next agent. Slack-style three-pane UI, JSONL event log, zero dependencies.

Open coordination protocol for AI agent + human teams. Inspired by agenthub.

灵通 (LingFlow) — 多智能体协作工作流引擎 | Multi-agent collaboration workflow engine

AI coding agent orchestration - run multiple terminal sessions with scoped context, todo lists, and parent-worker coordination

TO BE CEO : An AI boardroom that pressure-tests startup ideas through research, debate, scoring, and replanning.成为CEO,把创业想法丢进一家 AI 公司里跑一遍：研究、部门评审、董事会结论和重规划建议。

Local-first AI agent workspace with visual runtime, HITL control, sandbox audit, memory, RAG and skills.

AI copilot for React apps. Reads your state, fills forms, confirms actions, calls your tools. Multi-agent, AI SDK 6 + AG-UI, markdown skills. MIT.

Command a swarm of LLM-piloted drones in plain language: a Commander agent dispatches tasks to per-drone Claude agents that fly PX4/Gazebo SITL and report back, viewed live in a web Observatory. Self-contained in one Docker container (Gazebo + PX4 + ROS 2).

Self-hosted GraphRAG + multi-agent social-media simulator. Upload a doc → build a knowledge graph → spawn persona-driven agents → generate a report. Ollama & Neo4j instead of Cloud-APIs. DACH-first. Lokaler Multi-Agent-Social-Media-Simulator mit GraphRAG-Pipeline. Dokument hoch, Knowledge-Graph bauen, Personas spawnen, Report raus. Ollama + Neo4j

AI coding autopilot for PRD-driven software delivery with multi-agent planning, real pytest, peer review, and no-fake-run guarantees.

Enterprise Supply Chain AI Assistant powered by Amazon Bedrock AgentCore & Strands Framework. Features secure VPC isolation, OpenSearch RAG, and automated CDK deployment.

Lingtai-powered roundtable workflow for multi-perspective AI review: temporary experts, disagreement maps, accountable execution

Multi-chat orchestration plugin for Claude Code. File-based inbox, peer messaging, cross-session search across all your chats.

中文原生 Claude Code 配置平台 — 19 Agent | 按需注入 | 分层可插拔 | 零密钥上手

诊所 Clinic: a Codex/Claude skill for multi-agent consultation, persona reasoning, and live evals.

core orchestration engine for oneticket V1

🔱 Atlantis — the orchestrator-city of agents for Claude Code. One petition → expert Artisans (parallel, isolated) → Guardians audit → the three Judges weigh each blocker → one Decree. Config-driven, single file, zero deps.

DAG-scheduled multi-session epics with parallel-worktree execution. Works with Claude Code and OpenCode.

Local Multi-Agent Orchestration

A lightweight, single-file OpenClaw dashboard for real-time agent fleet monitoring — no build step, no database, no Docker.

A股量化智能体 - 10个AI分析师协同决策，LangGraph多智能体工作流，本地模型零成本运行

🌐 Automate your web tasks with Nanobrowser, an open-source AI tool that integrates seamlessly into your browser for a smarter browsing experience.

UNO card game with AI agents trained via PPO reinforcement learning, transformer architecture, and real-time Pygame visualization

AI 职场生存模拟与策略训练系统：卡牌博弈、多 Agent 决策、经营数据分析与执行反馈闭环。

13 autonomous AI deities running a spec-driven development pipeline inside pi - from vague idea to shipped release.

SNAP - Signed Network Agent Protocol. Secure agent-to-agent (A2A) communication with cryptographic identity, built on Bitcoin Schnorr signatures and Nostr.