Custom agents

XMPro's repository focusing on generative AI agents and Multi-Agents.

[KDD 2023] Ball Trajectory Inference from Multi-Agent Sports Contexts Using Set Transformer and Hierarchical Bi-LSTM

Craftsman ecosystem for AI coding agents (Claude Code, Hermes Agent, ...): blocks the write when it breaks your architecture, same rules in hooks and CI. DDD, Clean Architecture, TDD. Apache 2.0.

Extract documentation for AI agents from any site with llms.txt support. Features MCP server, REST API, batch processing, and multiple export formats.

Plugboard is an event driven modelling and orchestration framework in Python for simulating and driving complex processes with many interconnected stateful components.

Learn about the fundamentals of LangGraph through building an ambient email agent

We introduce ADAM, An emboDied causal Agent in Minecraft, that can autonomously navigate the open world, perceive multimodal contexts, learn causal world knowledge, and tackle complex tasks through lifelong learning.

SWE-Debate: Competitive Multi-Agent Debate for Software Issue Resolution [ICSE 2026]

Claude Code plugin that enforces hexagonal architecture (ports & adapters) in Symfony projects — with 10 auto-triggered skills, 2 review agents, and progressive refactoring support for existing codebases.

Agent-native voice + vision OS for wearables. Voice-first, agent-first framework for smart glasses, earbuds, and beyond.

AI-powered SOC for OT/ICS networks — 6 autonomous agents, MITRE ATT&CK mapping, Suricata/Zeek ingestion, human-in-the-loop response, 330+ tests. Open-source alternative to Claroty/Dragos for SMBs.

CLAUDE.md + AGENTS.md 12-rule behavior file for Claude Code, Codex, Cursor, Hermes — drop-in, ~700 tokens, cuts AI coding mistakes ~40%→~3%

PocketFlow from 0 to 1 ｜ 100 行代码构建所有 LLM 应用 ｜ 首个 PocketFlow 交互式教程 ｜ 光速掌握智能体开发实战

Customizable Laravel 13 starter kit for stateful/stateless apps with full code ownership and AI-ready architecture.

Custom n8n node for integrating Langfuse with AI agent workflows. Provides logging and tracing support for LLM-powered automation in n8n.

An intelligent novel writing agent for consistent long-form storytelling. Powered by LLMs.一个智能小说写作 Agent，专为维护长篇故事连贯性而设计。

Teneo Protocol AI agent skills for Claude Code and AI coding assistants

Official Implementation of ICLR2025 Paper: Songyuan Zhang, Oswin So, Mitchell Black, Chuchu Fan: "Discrete GCBF Proximal Policy Optimization for Multi-agent Safe Optimal Control".

🚀 Enterprise-grade API Gateway for MCP Protocol, built with Java Spring Boot. Supports authentication, proxy, and traffic management for AI tool calling.

A sandboxed execution environment for AI agents via WASM

Agent Sven: Terminal AI Agent

Operational patterns for AI agents and infrastructure from solo to enterprise scale. Applies DevOps + SRE principles to the intersection: infrastructure FOR AI + AI FOR infrastructure. Alpha framework extracting meta-patterns from mission critical environments.

LLM-agents benchmark set of geospatial tasks requiring multi-step tool use; and LLM-as-Judge based evaluation framework.

A distributed platform for AI Agents. Supports multiple LLMs, orchestration/subagents, long term memory, dynamic UI and more.

A Zed Agent Client Protocol (ACP) Jupyter Kernel

Secure autonomous AI agents in 60 seconds.

FinMTM: A Multi-Turn Multimodal Benchmark for Financial Reasoning and Agent Evaluation

AI Marketing Campaign Post Generator is an **AI-powered marketing campaign generator** that transforms business ideas into professional marketing campaigns using Google's Advanced Development Kit (ADK) Framework and Gemini API.

🎯Awesome-AgenticRAG_DeepResearch: A curated list of resources on Agentic RAG & DeepResearch. 学习参考关于AgenticRAG、DeepResearch的发展相关论文

Open Source Solution for Claude Creative Work

Awesome all-in-one research skills for AI agents — covering every academic discipline. Works with Claude Code, Codex, Gemini CLI & Cursor.

Prover Agent: An Agent-Based Framework for Formal Mathematical Proofs

Codebase readiness scoring for autonomous agents — measurable operability standards beyond instructions.

SpecMem: 🧠 Unified Agent Experience and Cognitive Memory for Every Coding Agent

企业级的数字员工容器化管理平台

AI 全链路小说创作工具

An open-source, AI-driven Social Media Digital Twin powered by LLMs for Computational Social Science simulations, network analysis, and human-agent interaction.

A multimodal RAG dashboard and interactive 3D Knowledge Graph. Process documents locally with Ollama or via Cloud APIs, powered by LightRAG, RAG-Anything and Neo4j.

🚀 多Agent协作智能知识库系统 — 五层可编排Agent架构 + RAG全链路 + 统一ToolRegistry，Spring Boot + FastAPI + React，核心亮点是 五层可编排 Agent 架构——Interface、Orchestrator、Tool、Memory、Evaluation 五层解耦，支持意图识别→问题改写→检索→充分性判断→答案生成的完整工作流。系统内置 Router / Retrieval / Ops / Inspection 多 Agent 协作，通过统一 Tool Registry 管理 AI 能力，支持 runId/traceId 全链路追踪、四级记忆体系与 Caffeine + Redis 多级缓存。适合Agent后端开发学习与简历展示

An opinionated Obsidian second-brain template, built for long-term human-AI collaboration

All your projects in one workspace. Run multiple CLI agents in parallel across different repos. Runs on your hardware, accessed through the desktop app or the browser. Bring your own plans.

Whole-codebase knowledge for AI coding agents. A field-aware code graph (functions, classes, methods, fields, references) plus persistent memory. Rust, Postgres + pgvector, MCP.

AI-powered development tools. 19 agents, 22 commands, 32 skills, 1 hook, 1 MCP server for code review, research, design, and workflow automation.

A curated list of resources, examples, and tools for AGENTS.md — a simple, open format for guiding coding agents.

AI teaching Skill based on the Feynman Technique — makes YOU teach the AI to expose gaps in your understanding. Works with Claude Code, Cursor, Windsurf, and any LLM agent framework.

把高质量播客和长文 RSS 自动转成 karpathy-claude-wiki 兼容的 source-summary 页面 / Turn high-signal podcasts and long-form RSS into karpathy-claude-wiki compatible source-summary pages. Part of the zero-code AI investment research toolkit.

bhived is an MCP server that gives AI agents shared memory, skills, and tool discovery. install once, works in Claude Code, Cursor, and 15+ other agents.

24/7 AI 指挥官 + Agent 托管平台（Hive）。理解意图、规划任务、调度工具、记住一切。支持 Telegram/WhatsApp/钉钉/飞书/QQ/企业微信 多渠道。｜ 📖 配套书籍《从零到一造 Agent》已在微信读书上架（链接见⬇️）。

Local experimental RAG playground for exploring retrieval quality, corpus construction, confidence and filter-chain design. Configurable retrieval orchestration, ranking, filtering pipelines, visual document grounding, web search, Open WebUI integration, and rich debugging insights. Ollama and vLLM, Dev Containers or natively on Win and Linux.

[ICML 2026] This is the official implementation for paper HiPER: Hierarchical Reinforcement Learning with Explicit Credit Assignment for Large Language Model Agents

电商全域数据治理与 Multi-Agent 分析中台：基于 n8n + LLM 架构，实现多源电商数据 ETL、规范化数仓分层建模 (DWD/ADS) 及飞书端智能对话查询闭环。

Autonomous SolidWorks AI assistant — embeds Claude Code / Codex CLI in the Task Pane to automate feature operations via natural language.

RoR engine that enables natural language queries over your database using RAG and LLMs.

Use Cursor CLI from Claude Code to delegate coding tasks to Composer 2.5+ and other Cursor models.

MCP server for real-time world news — free, no API key required

A trusted data asset platform for government and organizational collaboration, covering the full lifecycle of data governance, trading, authorization, analytics, and AI-powered exploration.

Index, search, summarize, and approval-gate local Codex sessions through OpenClaw/MCP. Agent-first recall and dry-run orchestration for Codex-heavy work.

SEO Content Engine for LLMs - a prompt-engineering system that turns any capable AI into a world-class SEO strategist: generate, rewrite, and audit content against 49 ranking factors with multilingual support.

Desktop coding-agent interface powered by Forgecode

Deterministic agentic-style multi-hop RAG at research-SOTA parity on commodity LLM APIs — no GPU, proof tree per answer.

Hermes Agent proficiencies: shareable operating-attribute skills for human SOPs/playbooks, agent skill/tool-SOP building, safe publishing, and tidy workspace hygiene.

Official OpenClaw Skill for Vorim AI — cryptographic identity, scoped permissions, and tamper-evident audit trails for OpenClaw agents.

Agentic AI development loop that splits Design, Code, Review, and Test into isolated roles with deterministic guardrails

Open-source, agentic web testing platform.

The Agno demo showcase — agents, teams, and workflows built with Agno.

Browser extension with 35+ rules for keeping your AI agent safe while browsing

A complete voice AI building block with telephony integration, featuring real-time speech-to-text, text-to-speech, and LLM-powered conversational agents.

探索 Claude Agent SDK 能力的实验项目。通过 Skill 系统让 AI 理解 SDK 文档，进而自主构建了一个 TikHub 社交媒体数据对话助手。 本项目是利用 .claude/skills/claude-agent-sdk 建立 claude-agent-sdk 项目， 项目基于 https://github.com/liangdabiao/tikhub_api_skill tikhub skill 封装成 webapp。

Tamper-evident evidence and verification layer for AI agents. Verify any agent's session offline with just openssl. No cloud, no lock-in.

A CLI-based multi-database tool that exposes database connections, query execution, metadata inspection, and connection reuse as local commands callable by agents. 基于 CLI 的多数据库操作工具，将常见数据库连接、查询、元信息读取和连接复用能力封装为 Agent 可调用的本地命令。

在 MiMoCode 的 Agent 环境下，纯文本模型 MiMo-V2.5-Pro 通过智能体编排获得多模态能力：MiMoCode 按需调用具备原生视觉能力的 MiMo-V2.5 处理图像输入，集成 MiMo-V2.5-ASR 实现语音指令转译，并将多模态上下文结构化存储于三重记忆系统中。纯文本"大脑"负责深度推理，多模态模块负责感知输入，协同完成从语音/图像到代码的全流程闭环。

EdgeGlow 是一个开源的 macOS 菜单栏小工具，让 AI 编程过程「可见」。当 Claude Code 或 Hermes Agent 工作时，屏幕边缘会亮起流光效果；完成后流光淡出消失。通过 HTTP hooks 与 AI Agent 联动，采用四层 CAShapeLayer 模拟霓虹灯光效，Timer 驱动动画确保流畅稳定。仅 892KB，~0% CPU 占用，支持多显示器和多终端。

Agentifying Patient Dynamics within LLMs through Interacting with Clinical World Model

Collection of Agents and Skills for Magento2

FAIR cyber risk quantification toolkits by Laura Voicu: agent-based control simulation (FAIR-CAM), threat event frequency estimator (PyPI), LLM classification validator (PyPI), Monte Carlo risk engine with IRIS benchmarks.

Autonomous AI infrastructure monitoring using Hermes Agent. Cron-based research ingestion, cost forecasting, and headless systemd deployment.

Reusable Agent Skills for Claude - technical-PM, prd-review, technical-pm-guided. Built by Taya Intelligence (Tigermed).

A multi-tenant vector database focused on extreme RAM efficiency. Lightweight, scalable, and optimized for high-density deployments.

OpenCode plugins that help AI agents understand their identity across multi-agent sessions

基于 MAPPO (Multi-Agent Proximal Policy Optimization) 深度强化学习的多无人机协同覆盖路径规划与动态避障算法

ICRTL Benchmark: Industrial-level RTL design challenges for evaluating PPA optimization, code generation, and LLM applications in EDA.

local RAG extension for pi coding agent

Benchmark Everything Everywhere All at Once, a fully autonomous agentic system for benchmark construction and customization.

Yak is an autonomous coding agent for papercuts. It picks up small tasks from Slack, Linear, Sentry, and GitHub and delivers reviewable pull requests while you work on what matters.

A multi-agent orchestration skill for long-running tasks.

Multi-agent swarm orchestration for Laravel, built on Laravel AI.

A skills & discipline framework for Microsoft 365 Copilot Cowork — for knowledge workers using AI as a coworker.

Local-first markdown notes with a first-class MCP server. 16 tools + 2 resources for Claude Code, Claude Desktop, Cursor and any MCP agent. Per-folder permissions, semantic RAG via LanceDB, single Bun binary. AGPL.

An open-source AI workspace for knowledge workers. Bring your own AI subscriptions and use the providers, harnesses, and models you choose to research, write, code, manage files, automate tasks, and control your computer and browser—all in one app. Built to respect and adapt to your workflow.

Automates chat replies for Messenger and WhatsApp using PyAutoGUI for UI actions, Pyperclip for clipboard capture, and Groq for AI responses. Cycles configured chat positions, skips replying to your own messages, supports dry-run and logging, and includes cursor tools and easy config for fast testing and demos. Ideal for portfolio demos & research.

Fable 指揮官工作流 — 最強模型只做規劃/審查，研究與執行派給你選的模型；maker/verifier 分離的執行迴圈。Commander workflow for Claude Code: plan & review on the top model, delegate research/execution, based on Loop Engineering. 中英雙語。

Official onboarding skill for HWPX document automation with AI agents.

AI-powered study assistant platform with quiz generation, document analysis, progress tracking, and smart learning tools built using Next.js, TypeScript, Prisma, Supabase, Gemini AI, and Redis.

TwinCAT coding agent for Beckhoff PLC development, with strict rules for Structured Text, naming conventions, PLC-aware tooling, syntax checks, and safe project validation workflows.

A runtime-agnostic local agent workflow that turns video links or transcripts into neutral, readable Obsidian notes — captions or local Whisper, free and fully local.

A local-first studio for roleplay and long-form writing: a roleplay chat with a living Narrative State, a Worldbuild that keeps your canon, and a Writing Desk for your chapters. Bring your own key or run local models. Free and open source.

Open-source test harness for AI agents. Stress-test production agents with adversarial multi-turn scenarios in CI

Concurrent Go CLI that detects hidden prompt-injection content in documents — invisible Unicode, homoglyphs, hidden HTML, and known trigger phrases — before they reach a RAG pipeline.

Demo: AI-driven autonomous research from synthetic data to peer-reviewed manuscript. Part of SciTeX.