Custom agents

Port of pi-mono to Python; AI agent toolkit: coding agent CLI, unified LLM API, TUI & web UI libraries, Slack bot, vLLM pod.

HiPRAG (Hierarchical Process Rewards for Efficient Agentic Retrieval Augmented Generation) is a reinforcement learning method designed for training reasoning-and-searching interleaved LLMs with improved efficiency and reduced oversearching as well as undersearching behavior.

An OpenClaw plugin that gives your AI agent a background reflection process and nightly dream cycle — encrypted memory, autonomous synthesis, and surreal narrative generation.

[IEEE T-ITS] A curated list of awesome resources on collaborative (cooperative) perception: papers, datasets, code, and benchmarks for multi-agent and V2X perception in autonomous driving.

Multi-agent chatbot with integrated search engine, utilizing a plan-execute-replan approach for complex queries, built with LangGraph and Tavily.

Self-adapting, Tool-enabled, Extensible, Multi-agent System

AEGIS: Automated Error Generation and Attribution for Multi-Agent Systems

Claude Code dotfiles: agents, skills, commands & includes for AI-assisted development. Features domain-specific agents (backend, frontend, DevOps, security, data/ML), custom slash commands, and Opus 4.5 optimizations.

MCP server for executing CMD commands. Can be hooked to claude for additional agentics.

Real-time dashboard for browsing, inspecting, and chatting with Claude Code agent sessions. Desktop & browser.

"Run AI coding agents (Claude, Mistral, other providers coming soon) in parallel inside hardened, isolated containers with git worktree separation.

Wiki-based retrieval for AI coding agents. 65× token reduction, +24pp Coverage@5 on SWE-bench Verified.

FylloCode 是一个开源桌面应用，把 Claude Code、Codex 等本地 Coding Agent 接进项目工作流。派任务、审方案、追溯每一次改动，积累下来的规范和结论，换一个 Agent 也能继承。

Just-in-time API keys for AI agents - and any other process you route through it: the caller only ever sees a placeholder.

Windows 本地/局域网 AI 助手：Web 对话、多用户与角色权限、团队文件工作区。Python 3.12 一键部署，依赖包已含（Git LFS）。

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.

Low-token CLI for AI agents to control a live Unity Editor.

🪢 Local-first, agent-agnostic R&D harness — persistent memory, domain SOPs, contract handoffs & quality gates for PM + Engineering | 本地优先、工具无关的 AI 研发治理与交付框架

All-in-one open-source framework & middleware for enterprise-grade multi-tenant and multi-tier XaaS Services development

SDK for Android — gives local LLMs structured access to phone data (calendar, contacts, SMS, files). Everything stays on device.

OpenATP is an open-source Python package providing a common interface for Automated Theorem Proving (ATP)

Agent-driven design studio for ESPHome-compatible (e.g. ESP32/ESP8266, RP2040) + LoRaWAN devices: generates ESPHome YAML, KiCad schematic/PCB, 3D-printable enclosures, and builds/flashes/provisions LoRaWAN firmware (ChirpStack + Home Assistant).

A persistent cognitive workspace that manages information flow for knowledge workers

The agent harness you steer — structured context, automatic staleness gates, and block-by-block review, all bound to your code. Local-first, MCP-native, open source.

Convert PDF, Word, PowerPoint, HTML, email & 40+ formats to clean Markdown — and back. Built for LLMs, RAG & Python pipelines, with a built-in MCP server.

Semark converts PDFs, Office files, HTML, and images into RAG-ready semantic Markdown using MinerU and VLM/LLM review, with chunks, source maps, and automatic document splitting.

把公司 Agent 化:按你的组织架构给每个部门(业务/财务/人力/行政/运营/合规)嵌一个 AI agent,CEO 参谋长统筹全局经营,常驻飞书。 | Agentify your whole company: an AI agent per department, a chief-of-staff running the business, all in Feishu/Lark.

Run DeepSeek as a real sub-agent inside Claude Code / Codex CLI — DeepSeek gets its own 7-tool agent loop in a sandboxed workspace, not just a single LLM call.

(New) The AI bookmark manager for busy people. With support for links, notes, images, and video previews from your favorite platforms

Aurict is an open-source terminal AI coding agent built for real developer workflows: sandboxed tools, subagents, multi-provider LLMs, patch safety, and fast CLI UX.

Evidence-centered academic literature workflow for topic definition, search, PDF download routing, Zotero management, paper writing, citation audit, and polishing. 面向中文/双语论文写作的证据闭环工作流。

An agentic AI tool for searching, screening, and ranking NIH funding opportunities based on research interests and proposal relevance.

Uxnan (pronounced /uʃ.nan/) is a suite of tools I built to solve a very specific problem I face as a developer: managing AI-powered coding agents from anywhere, without my hardware becoming a bottleneck.

Use WeChat to connect Claude Code

Unlimited context for Ai — give any LLM a billion-token memory. Open-source, local-first context + memory engine for agents. 🧠⚡

Framework for scientific llm-based multi-agent systems

A stack of tools and agent specialists built to help those who use LLMs as healthcare advisors.

Event-driven long-horizon memory for real-time wearable AI and LLM agents.

An OWASP-aligned intentionally vulnerable platform for learning and testing AI, LLM, RAG, MCP, and Agentic AI security.

基于 LLM 的保险保全智能助手 Demo，集成了 RAG 检索增强生成、Tool/Function Calling、多轮对话记忆管理、意图识别与路由 等核心技术，实现智能问答、自动化客服、业务 Agent 等典型 AI 应用场景。本项目以保险保全真实业务场景为案例，涵盖退保、手机号变更等核心保全业务，展示了 AI 技术在传统金融保险领域的落地实践。

An agentic system for autonomous neuroimaging workflows

Agent skills for autonomous software engineering, with focus on AI/ML

Autonomous multi-agent orchestration engine for software repos — L0/L1/L2 agents, leases, gates, audits, git-worktree isolation. Detached dispatch by default. GPL-3.0.

MDrive: Benchmarking Closed-Loop Cooperative Driving for End-to-End Multi-agent Systems

基于Multi-Agent架构的智能旅游助手

Build, deploy, and operate AWS infrastructure using AI agents, skills, and safety guardrails

Agent-to-agent messaging for Claude Code sessions on the same machine.

AI 视频脚本创作助手 通过自然语言对话生成专业 Seedance 2.0 视频脚本和分镜提示词的 Web 应用。基于 Claude Agent SDK 构建，内置完整的视频脚本创作技能，支持从故事分析到分镜脚本的全流程 AI 创作。

A minimal AI coding agent in 1,646 lines of TypeScript — the essence of Claude Code, distilled

Claude Code skill that prepares Indian Income Tax Returns (ITR-1/2/3/4): parses Form 16/AIS/26AS/broker statements, reconciles every rupee, compares regimes, computes tax + interest, and emits a portal-ready data-pack JSON for a browser agent to file.

Academic advisor due-diligence skill for Codex, Claude Code, and other agents | 学术导师调查评估

MM-ToolSandBox: A Unified Framework for Evaluating Visual Tool-Calling Agents

Ecommerce Agent 是一个基于大语言模型的垂直电商客服智能助手系统，专门针对电商客服场景进行深度优化。项目通过 LoRA 微调、DPO 对齐和 ReAct 推理框架，构建了一个能够理解复杂业务逻辑、合规调用工具、并提供专业客服服务的智能助手。

arcade/ 游戏大厅 A lightweight framework for text-based simulator games played by AI agents. Share energy, gold, and trophies across multiple games. Drop in a game directory, write a cmd(text) function — done. 一个给 AI 玩的文字模拟器游戏框架。跨游戏共享精力、金币、奖杯系统。把游戏放进目录，写一个 cmd(text) 函数，就接入了。

Inkwell is a production-ready "agent factory" built on the Microsoft Agent Framework: teams can deploy it for internal collaborative use, and individuals or OPCs (One Person Companies) can self-host it for personal use.

A command-line interface for Garmin Connect - activities, health, workouts, devices, and more - with JSON output, secure credential storage, and a bundled agent skill for AI assistants like Claude Code, Codex, Hermes, and OpenClaw.

Two agents spar over one codebase until every review finding is reconciled.

一个纯前端的 AI 角色语音聊天网页，5 个模型版本可选（Claude/DeepSeek/Gemini/GPT/Grok）

everything about llm based agent

Reproducing AgenticQwen (arXiv:2604.21590) — dual-flywheel data synthesis + GRPO RL training for agentic small LLMs

Pi coding agent extension for LSP integration — gives the LLM access to diagnostics, hover, go-to-definition, references, symbols, and rename

TRACE is a practical agentic jailbreak framework that leverages task decomposition, task-aware disguising scenarios and adaptive self-evolution to investigate emerging security risks of LLM-based agents.

Visual multi-agent orchestrator with a drag-and-drop DAG canvas, dynamic LLM planning loops, permission-based tool execution, and local Qdrant RAG.

CORTEX is an AI Agent framework designed for efficient integration and utilization of large language models (LLMs), built in Go.

现代化的AI驱动知识管理与智能体工作流平台，集成了大语言模型(LLM)、Skills、向量数据库、智能体(Agents)、工具集成，MCP，A2A协议等功能。该平台旨在帮助个人和团队高效地组织、存储、检索和利用知识资产，同时通过可定制的工作流自动化复杂任务。

Autonomous YouTube Shorts automation system built with n8n, AI agents, and FFmpeg that discovers viral trends, generates scripts, creates videos, and uploads content automatically.

Sirvir — model fleet manager & intelligence engine for Hermes Agent. Autonomous model lifecycle manager: local serving, benchmarking, auto-scaling, API fallback, HuggingFace scanning, creator tracking, backend optimization, budget monitoring.

Transaction Tokens (RFC 8693) for Kubernetes — seal identity, context, and authorization across multi-hop agent workflows via AgentGateway + ext_authz.

这是一个最小化的Multi Agent项目模板，帮助你理解 LangGraph 1.0的核心概念，并快速上手开发自己的多Agent应用。

A persistent management layer for Claude Code Agent Teams

Build, deploy, and orchestrate teams of Claude Code AI agents on Slack

The first MCP (Model Context Protocol) server for NetLogo, enabling AI assistants like Claude to create, run, and analyze agent-based models through natural conversation.

Claude agent skill: Create AI talking-head videos with 1,500+ avatars. Script writing frameworks + Creatify API automation.

such framework. very agent. much deploy. Агент ведёт фичу от идеи до прода. Человек занимается инженерией, а остальное сделает автоматика

VS 2022 extension that teaches GitHub Copilot how to write X++ for Dynamics 365 Finance & Operations

agentic AI coding assistant supporting claude code, ChatGPT codex, OpenCode and Pi coding agent; with API provider OpenRouter Gemini DeepSeek

MetaTrader 5 in a real Windows VM (Docker + QEMU/KVM), driven over REST + MCP — programmatic trading plus server-side TA: OHLC bars enriched with RSI/MACD/Bollinger/ADX/VWAP/Ichimoku/Order Blocks/FVGs (primitives, not signals). Typed MCP tools mirror the whole surface. Multi-broker, multi-account. No Wine bullshit.

Native Excel (.xlsx) generator tool for Open WebUI — multi-sheet, Tables, live formulas, charts. MIT.

A short-term planning system for engineering teams — keep engineers focused, run daily sprints, and track unplanned work. Built on GitHub Projects v2 (no database of its own): one Go binary with an embedded React UI, REST API, MCP server for AI agents, and live board updates over WebSocket.

An educational macOS app: build and chat with a tiny language model, trained on your Mac

AWS-native knowledge graph RAG framework unifying two graph-retrieval methodologies — Microsoft GraphRAG community summarization and LightRAG dual-level keyword search — on Amazon Bedrock, Neptune, and OpenSearch. Multi-hop QA over document corpora, with incremental indexing and multilingual support.

🕸️ Engineer the organization, not just the agent. 562 curated resources · 9 design layers · 11 sections · 252 papers & preprints — a field guide, CC0 open dataset, and interactive atlas for graph-structured multi-agent systems: roles, topologies, handoffs, work graphs, state, gates, reliability, observability.

Build your own world with AI agents inside Spherse

hold your voice helps writers keep their voice when they work with ai.

Local browser agent: MLX + Chrome DevTools Protocol. Handles cross-origin iframes, Shadow DOM, ProseMirror. Runs on Apple Silicon, no cloud.

Claude Code scaffolding and first steps for complex brownfield projects

Argus is a desktop app that helps find and fix wasted agent work. Local, free and open source. Focuses on business tasks, not code. Works with Claude Cowork / Chat / Code, ChatGPT Work and Codex.

A market-driven Product Manager copilot (Agent Skills + Claude Code plugin) based on the Pragmatic Framework — guides PMs from idea to launch and produces a client-ready artifact (Markdown or .docx) at every stage.

Markdown 长文渲染成小红书 3:4 图片组的 Agent Skill，全文不删减，内置 5 种风格（Cursor / Claude Code / Codex 通用）

智简Copilot：面向大学生与青年求职者的 AI 求职工作台。简历编辑、JD 匹配、自荐信、投递看板、模拟面试与复盘报告一站完成。

🐾 A desktop pet that reacts in real time to your AI agents — Claude Code, Codex, Antigravity, OpenCode, Cursor, Copilot, Gemini.

面向小红书/RedNote 创作者的 Agent Skill：图文轮播规划、视觉导演、封面内页设计、AI 生图提示词与发布文案。支持 Cursor、Claude、Codex。

🚀 Java AI工程师成长路线图 | 从Java基础到AI应用开发全栈学习资源 | Spring AI · RAG · Agent · MCP · LangGraph4j · 面试题库

Your AI agent. One server. Any device. No baked-in behavior.

安卓原生 bilibili 客户端，提供无推荐流的订阅式设计、优化的听视频功能和 Agentic 搜索推荐能力。

Turn research and long-form writing into humanistic, model-interpreted Xiaohongshu carousels.

See your AI agents work. Real-time 3D visualization of Claude Code activity, multi-agent orchestration, and REST API integration for external systems.

Use DeepSeek V4 Flash inside the stock ChatGPT desktop app / Codex model picker — native Responses API, full agentic tool loop, no fork, no app patch. GPT models keep working via your ChatGPT OAuth subscription.