Custom agents

A tool for fast and accurate summarizing of variant calling format (VCF) files

INDUSTRY ALERT: The Subprime Code Crisis. A data-driven analysis of how AI assistants inflate a bubble of technical debt. Featuring research on 211M lines of code, this report exposes the "Placebo Analytics" of AI productivity and provides a framework for engineering governance.

A unified python SDK supports OceanBase or OceanBase seekdb, more efficient and easy-to-use.

Neural Language Model for Go

An AI-powered assistant that simplifies DevOps tasks with intelligent prompts, reducing repetitive work and enhancing productivity

A Go library implementation of LightRAG - an advanced Retrieval-Augmented Generation (RAG) system that uniquely combines vector databases with graph database relationships to enhance knowledge retrieval.

SpikeAgent is a multimodal LLM-based AI agent that automates and standardizes the spike sorting pipeline

NeurIPS'24 DB (Spotlight) | Instruction Tuning Large Language Models to Understand Electronic Health Records

General Tool-calling API Proxy

A Model Context Protocol (MCP) server implementation exposes document processing capabilities through natural language, supporting both direct human interaction and AI agent tool calling.

Powerful AI Assistant in Roam Research

MCP server for Claude Desktop to manage your n8n workflows

This is the official implementation of paper "Leveraging Dual Process Theory in Language Agent Framework for Simultaneous Human-AI Collaboration."

Fun LLM Agent Projects I Designed & Built

🐱 Pixel art cats on your macOS dock — debate ideas with local AI

FastAPI MCP server that exposes browser-use automation through a testable Model Context Protocol boundary.

TrustAgent: Towards Safe and Trustworthy LLM-based Agents

AIFA Network — open research on decentralized infrastructure for autonomous AI agents: Google A2A interoperability, Kademlia DHT discovery, NAT relays, end-to-end encryption, on-chain escrow and the $AIFA utility token. Specification stage.

AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents

Evaluate Multimodal LLMs as Embodied Agents

Clarity is a financial analysis agent framework built on native Claude-skill architecture. Adopting a Planning-with-Files approach, it coordinates specialized sub-agents via persistent task plans, research findings, and progress logs to tackle complex financial analysis workflows.

Code for "TrustRAG: Enhancing Robustness and Trustworthiness in RAG" AAAI 2026 Workshop on Trust and Control in Agentic AI (TrustAgent)

A Model Context Protocol (MCP) server that enables AI assistants to interact with Kubernetes clusters. It serves as a bridge between AI tools (like Claude, Cursor, and GitHub Copilot) and Kubernetes

Datadog for AI Agents.

Stop re-explaining your research to your AI agent. Persistent, LLM-maintained wikis that compound over time. Drop PDFs, URLs, YouTube - your agent remembers forever. Based on Karpathy's LLM Wiki pattern.

Tara Bot — AI Telegram agent: search flights, compare prices, inject affiliate links

AI knowledge assistant using multi-agent processing with context-aware analysis

## AI医院: Multi-Agent 医学智能体 与患者进行多轮对话，提出相关和探索性的问题，推荐适当的医学检查，并在收集足够的信息后做出诊断。设置检查员，他们专门负责与患者互动并提供相关的医学检查结果，确保医生能够获得患者必要的客观信息以做出准确诊断。此外，主治医生负责在整个会话后评估医生的表现。 模拟患者与多名医生之间的互动，包括体检结果的收集和最终诊断。

MBA AI Agent课程：2天集中培训，从LLM基础到多智能体系统

Agent Skill. The only memory Claude Code needs exists in the file system.

Orquestrador de 5 agentes IA especializados para analise 360 de features com Claude Code. 1 comando, 5 perspectivas, 1 relatorio executivo.

A multi engine TTS & LLM edge computing playground with audio book features and more!

[NeurIPS 2025] The official implementation of the paper "DRIFT: Dynamic Rule-Based Defense with Injection Isolation for Securing LLM Agents".

Curated open-source skills for AI agents (Claude Code compatible)

A holistic framework for advancing LLMs as data science agents

Learn as you build. Build with intent. Ship real products. Agentic engineering done right.

A curated list of 260+ papers and resources on Agentic Security. Based on the survey covering the transition from passive LLMs to autonomous agents, structured around three pillars: Applications (Red/Blue Teaming), Threats (Injection, Poisoning), and Defenses (Guardrails, Verification).

WormGPT is an experimental AI project designed strictly for educational, cybersecurity research, and defensive testing purposes only. It is intended to help developers, researchers, and students understand AI behavior, prompt injection risks, and security vulnerabilities in language models.

Prompt framework that makes LLMs discover real, exploitable vulnerabilities. Six specialized agents covering recon, hunting, adversarial chaining, exploit validation, triage, and AI security.

Multi-agent AI trading system using LLM-powered analyst agents (Buffett, Munger, Burry, etc.) with free data sources (SEC EDGAR + yfinance) and Alpaca integration.

Production-grade design context for AI coding workflows. Extract a real design system from any URL — colors, typography, spacing, breakpoints — as a portable DESIGN.md.

A Rust framework for building AI agents using ECS-inspired architecture and graph-based execution.

A curated list of frameworks, tools, and resources for building and deploying AI agents.

Device Context Protocol — bridge LLM agents to physical devices. Sub-50-byte frames, 27.6KB flash / 0.6KB RAM measured on ESP32, capability-scoped and safe by design. Complementary to MCP. Paper: arXiv:2605.26159

An open-source Python SDK for analyzing, evaluating, and curating agent traces stored in BigQuery. Built on top of the BigQuery Agent Analytics, it provides a consumption-layer toolkit for agent observability, analysis, evaluation, and advanced capabilities like context graph at scale.

Complete collection of Claude Code configs for C++20 HPC development - agents, skills, hooks, commands, and rules

LLM-agnostic tabletop RPG Game Master framework. Runs on OpenCode, LM Studio, or any LLM service. D&D 5e included as reference system — add any TTRPG via a system module.

Multi-agent memory governance demo using MemClaw + OpenClaw. Three agents (Sales, Legal, Admin) share one memory backend with hard fleet boundaries enforced at the query layer. Shows scoped writes, blocked recall, cross-fleet synthesis, and conflict detection.

Matrix by PaxLabs is a production-grade framework for building autonomous AI agents that go beyond chat! Reasoning, using tools, writing and executing code, navigating files and the web, and securely carrying out real-world workflows, including payments, on-chain operations, and smart contracts, through natural language.

Local MCP extension for product discovery: personas, interviews, evidence analysis, PRDs and delivery briefs. Works with compatible agent hosts; bundled in Orqanix.

Agentic AI platform transforming passive CCTV into proactive railway safety intelligence. Real-time behavioural detection (suicide risk, pickpocketing) via YOLOv8 + BiLSTM + multi-agent reasoning — zero new hardware, privacy-first, no facial recognition.

SECOM: On Memory Construction and Retrieval for Personalized Conversational Agents, ICLR 2025

Personal usage analytics for Pi coding agent. Temporal-aware insights, context-aware suggestions, model efficiency analysis.

Baton — clean session handoffs: pass in-progress work to the next agent with a single verified markdown file. A Claude Code / Codex skill.

Agent Skills · 橙皮书系列 · Claude Code Skills 系统完全指南

A Kiro power that turns Kiro into a personal study companion for any AWS certification exam

An autonomous photonic device design agent iteratively optimizes silicon photonic devices — no human in the loop.

🤖 Autonomous Multi-Agent AI Workspace in your terminal. Codes, researches, and executes real tasks with 40+ tools and 280+ skills And Some Captcha Resolver With Only Vision Mode Any Captcha And Opening A Browser And Full Control Of It

Benchmark for evaluating LLM agents on smart-contract vulnerability discovery and exploitation

12 executable AI agent skills for competitive strategy analysis, extracted from Michael Porter's Competitive Strategy. Five forces, competitor profiling, strategic groups, market signals, generic strategies, competitive moves, and industry-type playbooks — grounded in Porter's original frameworks.

A desktop task manager for AI coding agents like Claude Code and Codex, with multi-project workspaces, live terminals, session tracking, permission-aware execution, Git integration, and usage analytics.

A collection of skills for agentic LLM tools

Natural language team builder for Claude Code Agent Teams — create purpose-driven AI teams with a single sentence

基于 LangGraph 和 FastAPI 的电商 AI Agent 框架。从 RAG 知识库到复杂的退货业务流的智能客服系统。

Trace-backed Agent Authorization Reviews for tool-using AI agents. Staging-only evidence for payment, record-change, access, and export actions.

本地化个人理财助手 — A股/基金/理财/现金/数字资产/机器人全资产看板 + 市场AI问答 + 个股K线盘口详情 + 板块对比 + 早盘信息简报 + 资讯AI解读。只给客观信息，不荐买卖。

AI Research Writing Skill (AI论文写作技能) is an agent skill for ML / AI / CV / NLP researchers. Point your coding agent at code, experiment logs, notes, and a venue template; it helps you produce an auditable, evidence-backed LaTeX draft and submission package — not a polished fiction.

Opinionated DevPod workspace configs for agentic engineering — Claude Code, shared secrets, VS Code extensions, ready on first launch.

A self-evolving harness framework for Hermes Agent. Evolution + Critic + Verifier + Gardener cooperate to make your Hermes instance measurably more useful over time. Generic, MIT-licensed, ships empty.

Official Repo for "AgenticASR: Refining Speech Recognition in Real-World Scenarios via an Agentic Approach"

Open-source BYOK AI coding agent with a desktop app, scriptable CLI, and built-in multi-agent orchestration. Written in Go.

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

Open-source AI funding workspace for consultants, advisors, and incubators. Manage opportunities, companies, applications, and AI-generated funding packages.

CLIProxyAPI 管理面板：查看运行状态、请求用量、费用、日志和配置，并执行带校验与回滚的安全更新。CLIProxyAPI dashboard for health, usage, logs, config, and safe updates.

Profanity-triggered self-audit protocol for coding agents (Claude Code, Codex CLI, Kimi CLI, OpenCode). Python stdlib only.

AI 驱动的自主渗透测试平台：输入目标，AI agent 自动完成侦察→规划→主动测试→漏洞验证→报告生成。黑板引擎+多 worker+SRC 验收门禁。仅供学术与安全研究使用。

FTShare financial data Skill for Claude Code, Codex, OpenClaw and AI agents｜FTShare 金融数据 Skill 接入方式。

Survey: The Path to Recursive Self-Improving Agents: Foundation, Framework, and Future Directions

A hierarchical, tree-structured coding agent engineered for weak and small LLMs to conquer complex software tasks via strict context isolation and interface contracts.

A chain for agents, on BNB Smart Chain. Genesis carries three system contracts and three neutral tools — Multicall3, a CREATE2 deployer and Wrapped BAC. Every token, pool and market beyond that is deployed by an agent. Agents never sleep, so they find each other's contracts in seconds. Pre-launch: nothing is deployed.

tmux agent monitor and session manager: live Claude Code, Codex, and Pi status plus a navigable session/window/pane tree in one sidebar/popup

A lightweight microVM sandbox runtime for agents, automation, and untrusted workloads.

Alice 集成技能包 · 放进去就能用 · Drop-in skill suite for AI coding agents

Zissa Wiki — an LLM-maintained research knowledge base in plain markdown, maintained by any AI agent (Claude, Codex, Gemini, DeepSeek, Kimi, Grok, Mistral or any chat app). Implements Karpathy's LLM Wiki pattern; checks every quote against its source.

AI software factory for the enterprise

Open protocol and TypeScript SDK for AI agent authority: identity, scoped delegation, enforcement before the action, and signed receipts.

Independent Android client for Codex with on-device tools, live voice, and Claude/Gemini collaboration. No Termux or PC server required.

An automated coding agent for small, cheap LLMs: split big goals into contract-driven task trees, execute each leaf in an isolated context, and verify with real builds and tests.

An AI video studio your coding agent installs and runs: UGC ads, micro-dramas, sitcoms, launch films, fantasy shorts, hour-long meditations. Claude Code / Codex + APIMart + Kie.ai + ElevenLabs.

One API and CLI over every coding agent (Claude Code, Codex, Copilot, Cursor, Gemini…), plus a thin provider-neutral LLM layer for TypeScript.”

Build a web app from a spec with Claude Code — feature by feature, through 18 checkpoints. By Open Reach Tech.

AgentDNS is the core component of Agent network that facilitates agent registration and discovery, enabling large language models (LLMs) to browse the internet freely, just like humans.

A curated list of awesome multi-agent learning papers

A Complete Open Source Agent AI Framework For Unity

Fact-Verified Travel AI Agent

A (really) easy way to web scrape

a self-hosted Linux + macOS agent that lets you remotely control your desktop via Telegram using natural language. Supports local AI for full privacy, bilingual parser (English + Arabic), smart multi-step recipes with state verification, and zero external dependencies.

A dystopia simulator powered by ChatGPT

New Relic CLI is a command line tool which is used to operate New Relic objects(Synthetic monitors, alert policies, conditions, account users etc). You can use it easily to list/get/create/delete these objects. It can be used to backup your New Relic configuration data and restore in the future. It is easy to be used other than calling different REST API endpoints.