Custom agents

TiDB AI SDK: Unified Multi-Modal Data Platform for AI Apps & Agents

Professional RAG development skills for Claude Code - audit, evaluate, optimize, and scaffold RAG pipelines

The repository contains the implementation Traffic Flow Optimisation for Lifelong Multi-Agent Path Finding. It plans and navigates more than 10,000 of agents in Lifelong MAPF problem,

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

把 CLAUDE.md、context/、.codebuddy/ 等 AI 配置文件统一放在一个 Git 仓库里，团队成员通过 clone 这个仓库获得完全一致的 AI "记忆"。

create_pg_super_document is a project that generates documentation for all symbols in the PostgreSQL codebase, then utilizes these symbol documents to create high-quality documentation for functionalities, components, and other elements using AI agents

这是一个基于AgentScope框架的AI助手应用，支持自动技能发现、智能任务匹配和终端交互，兼容claudecode的skill文件。

Local FAISS vector store as an MCP server – Agent Memory, drop-in local semantic search for Claude / Copilot / Agents.

Runtime visibility for Python MCP servers. Captures tool calls, session lifecycle, module imports (SHA-256), and subprocess execution as structured NDJSON. No code changes.

intelligent-virtual agent authoring toolkit with interoperability of rocketbox , cc4 and didimo characters

Open-source workflow automation platform with 120+ connectors, AI-powered workflow generation, and visual builder

End-to-end financial text-analysis using Bigdata API and the Bigdata-Research-Tools library. Ready-to-use notebooks with RAG & GenAI enabling thematic and risk screening, trend tracking, and automated report generation, extracting insights at scale.

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.

卡车-无人机协同配送系统，通过多智能体强化学习(MAPPO)算法优化物流配送效率。系统模拟真实的城市配送场景，其中卡车作为移动补给站，无人机执行最后一公里配送任务。optimized using Multi-Agent Reinforcement Learning (MAPPO) algorithms to enhance logistics delivery efficiency. The system simulates real-world urban delivery scenarios where trucks serve as mobile supply stations, and drones carry out the last-mile delivery tasks

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

This project leverages Claude Code’s powerful agent capabilities to build a multi-agent system that simulates the real-world collaboration process of a data science team. Through the division of labor among AI agents with different areas of expertise, the system delivers a complete end-to-end data science solution—from raw data to the final analysi

A Claude Code agent that analyzes external Claude setups and assimilates the best and most relevant features for your own. Resistance is futile.

This agent answers questions by reading multiple docs provided by the user

VLA^2: Empowering Vision-Language-Action Models with an Agentic Framework for Unseen Concept Manipulation

Terminal multiplexer with a first-class agent API for shared human+agent workflows

Persistent memory for AI agent teams: private silos, shared hive, and curator workflow.

Git-versioned backup tool for AI coding agent config folders (Claude Code, Cursor, Codex, etc.) with size-based rotation and easy restoration

A universal CLI for Weaviate, Milvus, Chroma, Qdrant, and other vector DBs to help view, list, create, delete, and search collections and documents in collections for development, test, and debugging purposes

Give your AI an operating system. Open-source framework for persistent AI context.

[AAAI'26, Oral] Modeling Uncertainty Trends for Timely Retrieval in Dynamic RAG

MCP server that turns raw sources into an interlinked Obsidian wiki with Claude Code.

Multi-Agent Solutions (Deep Research and more)

Multi-Agent Verification: Scaling Test-Time Compute with Multiple Verifiers

Official implementation of the paper "Learning and Planning Multi-Agent Tasks via a MoE-based World Model"

Agent-first FastAPI template built to pair with Claude Code - opinionated architecture kept on the rails by custom skills, hooks and review agents.

一个常驻桌面的 AI 小伙伴。悬浮球形态，轻量不打扰，支持对话、文件处理与轻记录，具备 Agent Loop 与按天记忆能力。Electron + TypeScript 构建，支持 macOS & Windows。

Mobile app for managing Codex, Claude Code, and soon Gemini on remote servers through a chat-style interface with terminal, diff viewer and file browser

Agent skills for building, shipping, and growing software products

A sophisticated Model Context Protocol (MCP) server for analyzing Architectural Decision Records (ADRs) and providing deep architectural insights to AI agents.

Open-source AI agent social network built with Next.js + Supabase. Self-hostable, cryptographically secure, API-first. MIT license.

A thin cython wrapper around llama.cpp, whisper.cpp and stable-diffusion.cpp

Debug-only in-app MCP server for iOS/macOS/Android/Flutter/ReactNative. Gives LLM agents Playwright-like control over native apps — UI, state, navigation, and diagnostics over the local network.

Universal messenger-to-agent bridge — connect WeChat/Feishu/Telegram to Claude Code/Codex/Copilot

Founder-first meta-layer to generate production-ready OpenClaw multi-agent setups from one prompt: agents, skills, workflows, memory, and install docs.

Papers about agentic AI and AI agents for spatio-temporal data.

Agent-Omit: Training Efficient LLM Agents for Adaptive Thought and Observation Omission via Reinforcement Learning

Starter project for the Build your first declarative agent for Microsoft 365 Copilot by using Visual Studio Code learn module

🦀 Mobile-first AI agent framework. Your butler, in your pocket. OpenClaw alternative for 6.8B phone users.

Skills that get AI assistants to stop agreeing with bad product ideas by default — pressure-test premises, separate hobbies from businesses, and refuse validation theater.

Qworld: Question-Specific Evaluation Criteria for LLMs

A shadcn/ui library for building ChatGPT Apps and MCP Apps

A token-economy playbook for GitHub Copilot Chat & agents under usage-based billing.

这是一个基于VSTO开发的Microsoft Office Word的AI对话插件，也兼容WPS Office

Open-source PDF-to-Markdown post-processor with footnotes, LaTeX normalization, figure links, and YAML metadata. Supports Marker, MinerU, PyMuPDF, and Docling. Includes a self-hosted web UI.

🎯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

Zero-config local LLM optimization for Ollama, LM Studio, and Apple Silicon MLX. Reduces TTFT by 40%, wall time for local agents by 46%, and RAM usage by 3x.

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

The goal of this project is to create a modular Agentic Python framework for robotics that can be used to develop specific robotic applications for various types of robots, IoT, and physical devices. The initial focus will be on the hospitality sector, with a longer-term focus on transitioning to consumer usage such as household tasks.

Turn a VPS into an AI-powered personal server with persistent memory, Telegram integration, and knowledge management

💎ᴠɪᴘ Free online Learning with Professional Certifications. Free online courses with certificates in various fields such as Gen AI, Prompt Engineering, Data Science, AI, ML. Welcome to our Online Course Portal project! 🚀

AI智能助理前端组件,聊天助手前端组件,AI组件

MCP server for Denmark's Aula school platform — TypeScript MitID auth, no headless browser. Exposes profiles, calendar, messages, ugeplaner to AI agents (Claude/Cursor/etc) via Model Context Protocol.

企业级的数字员工容器化管理平台

🚀 5,000+ Precision-Tagged AI Skills. Auto-updating & Agent-ready. / 5000+ 精准标注 AI 技能，自动更新，智能体必备。

《深入 OpenCode》。全网第一本介绍 OpenCode 的书，用 OpenCode + Opus 写成。A book introducing OpenCode, written by OpenCode + Opus

Agent Skill for PHP security audits - OWASP patterns, vulnerability detection | Claude Code compatible

A curated list of MCP servers, tools, protocols, and resources for AI-powered advertising campaign management. Covers Google Ads, Meta, LinkedIn, Reddit, TikTok, Amazon, and more.

Turns issues into merged PRs/MRs with zero human intervention — supports GitHub and GitLab, and Claude Code, Codex CLI, Kiro CLI, Antigravitiy CLI, opencode, and other agent CLIs.

A self-evolving memory layer for LLMs.

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

MCP server for Outlook email, calendar, and contacts — let your AI assistant manage your inbox directly from the conversation.

AI求职操作系统 — 岗位采集→10维评分→定制简历(LaTeX PDF)→面试资料生成→投递指南，一条龙服务

Model Context Protocol (MCP) server that gives LLM agents and chatbots direct access to the World Bank's Data360 Platform. Search, validate, and retrieve development indicators—from GDP and poverty to gender equality and climate—with structured metadata and time-series data, without hallucinating values.

🤖 LLM Engineering Roadmap — Complete Developer Guide

MCP server for Ansys Electronics Desktop automation with PyAEDT, HFSS, Maxwell, Q3D, Icepak, native AEDT APIs, reports, sweeps, and simulation workflows.

Borg Internet Operating System ( Built With PeerTree). Resistance is futile... The Internet became crap. Join the Borg and build something better

对话即创作 · AI Agent 驱动的小说写作工作台 ｜ Chat to create — an AI-agent-driven novel-writing workbench

Your coding agent starts every session blank. Rekal is the memory your team is missing

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.

MCP server that lets Claude Code delegate tasks to the Kimi CLI agent

QodeX — a local-first LLM agent & AI coding CLI agent for your terminal. Runs local models (Qwen3-Coder via Ollama/LM Studio) with deterministic guardrails, 100+ tools, a real browser, smart vision & shareable live artifacts. Open source, Apache-2.0.

AI agents now operate with authority. Authority without discipline is how complex systems fail. Nuclear’s control loop, ported to AI-assisted software engineering.

The infrastructure for the AI economy: primitives for inference, agents, and decisions.

A semantic event bus for AI-agent ecosystems

为 DeepSeek 前缀缓存定制的终端 Code Agent（纯 Go），缓存命中率 95-99%，输入成本降至 1/50。A terminal coding agent optimized for DeepSeek prefix caching — 95-99% cache hit, 1/50th the cost.

Unity AI 交互框架

Panacea is a framework for building collaborative, intelligent multi agent AI systems. The framework provides a robust infrastructure for creating and managing multiple AI agents, and enables developers and organizations to build, deploy, and optimize AI agents that work well in dynamic, complex environments.

Open-source, agentic web testing platform.

Agent-native disposable email — Cloudflare Workers + Durable Objects, with CLI & MCP. https://smails.dev

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

A constitution generator for autonomous, agent-run projects: an interview-driven meta system prompt (and installable skill) that scaffolds a CEO-orchestrated sub-agent + skill + CLI architecture.

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

MultiTable is an open-source AI agent framework and meta-harness that gives you a common orchestration layer over LLM providers

Open APA (Open Agentic Process Automation) is built on top of our first-place performance in the OSWorld Benchmark. It provides a practical and extensible framework for building computer use agents that can reliably execute long-horizon, real-world tasks across applications, interfaces, and workflows.

AI coding agent in your terminal. Works with any LLM - DeepSeek, Qwen, GPT-5, Claude, Gemini, Kimi, Ollama. 100+ models via litellm. ~1300 lines of Python.

[ICLR 2026] Stop Wasting Your Tokens: Towards Efficient Runtime Multi-Agent Systems

Language- and framework-agnostic template for agent-driven projects.

A collection of 5 practical AI projects that demonstrate how to build real-world AI applications using LLMs, agent workflows, tool calling, and memory

One registry, every agent. Browse and share skills across Claude Code, Cursor, Codex, and more.

[ECCV2026] Official Implementation of "VideoSearch-R1: Iterative Video Retrieval and Reasoning via Soft Query Refinement"

Memory Injection Attacks on LLM Agents via Query-Only Interaction

打破Claude对中国的歧视，做一个对接N个CLAW，并且实现多AGENT的可视化操作工具！