Custom agents

These are my vibecoded experiments (generated by the agents themselves based on the docs of various tools) -- many of the skills are not properly vetted and I am simply improving them as I go along and notice mistakes done by the agents as they use a given skill.

An agentic coding and automation assistant, supporting both local and cloud LLMs.

An Obsidian-first, agent-maintained LLM Wiki framework that turns raw sources into linked notes, concepts, summaries, and evolving schema rules.

Secure autonomous AI agents in 60 seconds.

🎯人人可用的Agent，你的终极数字奴隶。让Agent帮你处理一切，提前享受纯粹的未来生活，穿越到2050年的通行证 | A conversational AI agent capable of autonomous browser automation

Sketch2Simulation: Automating Flowsheet Generation via Multi Agent Large Language Models

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

Une suite complète de skills, agents et scripts marketing entièrement en français, conçue pour Claude Code.

Easy parallel project management for Claude Code agents.

Native Linux GUI for Claude Code agents. Rust + GTK4. Fast, light, no Electron.

Cursor AI Agent Team Framework: A three-agent collaboration system for AI-assisted work

Security-first AI agent daemon — the model proposes actions, the runtime decides what execute

Markdown-Native Multi-Agent Task Coordination

🔒 Jail your AI agents — sandboxed Docker environments with network isolation for Opencode agents

8v is one binary that reads, writes, searches, checks, builds, tests, and formats code, using fewer input and output tokens than the agent's native tools.

Plugins & tools to build self-sustaining & autonomous AI Agents on Aleph Cloud and LibertAI

Safactory: A Scalable Agentic Infrastructure for Training Trustworthy Autonomous Intelligence

Agent Skills for WordPress plugin development — security auditing, admin theming, license management, auto-updates, and more. Compatible with Claude Code, Cursor, VS Code Copilot, and all Agent Skills-compatible tools.

Web dashboard to run, monitor, and manage multiple Claude Code & Codex agents. Create agents with a cloneable task template. Real-time streaming, task pipelines, session resume, git worktree isolation, and remote access via relay — all from your browser.

An MCP server that provides AI assistants with screenshot capabilities — both web page capture via Puppeteer and cross-platform system screenshots using native OS tools.

🧠 SBTI 赛博人格测试 — 30 道题解锁 27 种沙雕人格，测自己或测你的 AI agent | The viral cyber personality test as a nexu Skill

Production-ready Playwright test automation starter for teams and startups: strict TypeScript, Page Object Model, environment management, interactive runner, custom HTML reporting (with optional JIRA integration), quality tooling, CI/CD patterns, and agentic (MCP-ready) extensibility.

AI 全链路小说创作工具

Database MCP server for MySQL, MariaDB, PostgreSQL, and SQLite - with builtin PII redaction and write-prevention

AEP (Agent Element Protocol) v2.8 | Deterministic zero-trust total control and governance protocol for AI agents. | Reduce hallucinations to zero through architecture in all constrained domains of application. | LLMs gave you the engines, AEP gives you the control thrusters.

Agent Workflow Kit — Evaluation-first rules, templates, and Skill packages for AI-assisted software projects. Score risk, choose the smallest safe workflow, and generate practical AGENTS.md guidance for any coding agent.

A modern, cross-platform functional-first shell programming

Open-source health AI agent platform for self-hosted health questions, personal health data, personalized Nutrition for Metabolic Health, AI meal planning, Telegram chat, and Apple Health sync.

Real, runnable Flue agents on Cloudflare. Examples teach one product at a time; recipes compose. Every snippet ships an E2E test that deploys an ephemeral Worker, asserts, and tears it down.

🚀 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

AI-powered multilingual FIR drafting assistant using Flask, NLP, and Machine Learning.

A structured workflow system for AI coding agents - harness engineering, execution loop, skills, hooks, and a learning loop. Works with Claude Code, Codex, Antigravity CLI, Copilot.

A Claude Code operating system for running a structured agency team. 25 specialized roles across four councils (Creative, Technical, Hardware, Delivery), enforced quality gates, PM-driven workflow, multi-client workspace governance, and per-project sprint tracking.

Postman for AI - design, evaluate, and debug LLM interactions with full transparency.

🤖 Визуальный конструктор Telegram ботов с drag-and-drop интерфейсом. Создавайте ботов без программирования, получайте готовый Python код для aiogram. React + TypeScript frontend, Express.js backend, PostgreSQL. Поддержка медиа, клавиатур, условной логики и аналитики пользователей.

基于本地 RAG 与大模型人格蒸馏的多平台聊天记忆系统。通过 FTS5 检索、语料清洗与长期报告生成，将聊天记录转化为高拟真的数字分身。

Curated archive of agentic AI samples, reference implementations, and tooling from the Agentic AI Acceleration program in Korea. AWS-native patterns covering Bedrock, AgentCore, and end-to-end agentic AI pipelines.

WaSP — WhatsApp Session Protocol. The missing infrastructure layer between WhatsApp and your application.

Build, deploy, and connect AI agents to SAP Joule via the A2A protocol on BTP Cloud Foundry — directly from Claude Code. Supports TypeScript (Express or CAP) and Python agents with LangGraph and SAP GenAI Hub.

HYBRO Hub: local & remote AI agents — private, powerful, unified.

Claude Code skill: Multi-agent adversarial review panel — 4-6 AI reviewers debate your code/plans, then a supreme judge delivers the verdict. 9 auto-detected signal groups, built-in domain checklists, anti-groupthink mechanisms.

Effect building blocks for agentic ai

🤖 Official Interactive Tutorial for OpenHarness – Zero to Hero in 12 Chapters | Learn OpenHarness like Claude Code: Agent Loop, Tools, Memory, Multi-Agent | 面向零基础的 AI Agent 交互式教程

A coding-agent CLI and strongly-typed Python library -- self-mutating, hot-swapping, multi-provider, with async tool calls and bidirectional recursive spawn.

WizTree + AI 智能磁盘清理 Agent skill

Autonomous AI decision engine for OpenClaw — give your AI a brain that decides for itself

Vim style TUI with Repository-Aware, Self-Evolving Agent That Understands Codebase in Real Time powered by Hybrid Semantic + Full-Text Engine

Unity AI 交互框架

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

A comprehensive collection of expert AI Agent Skills for Google Gemini, featuring a powerful CLI to convert and optimize Claude Code agents for the Gemini ecosystem.

Manage, install, update, and trace agent skills (Claude Code, Cursor, Codex CLI)

Build Your First Copilot Challenge (Foundry Edition)

Portable — a drop in replacement /remote-control on Claude, + batteries

Turn your codebase into readable format, for both Human and AI Agent.

Per-project semantic search over Claude Code session history. Indexes JSONL transcripts into SQLite; retrieves via bge-m3 embeddings (Ollama) with FTS5 fallback.

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。

COS-PLAY: Co-Evolving LLM Decision and Skill Bank Agents for Long-Horizon Game Play

A reinforcement learning framework with verifiable aesthetic rewards for improving aesthetic slide generation capabilities in LLM agents.

A secure agent harness runtime and framework

Reading copilot for Zotero powered by the DeepSeek API and scoped to your local papers, PDFs, and collections.

The first multi-level safety evaluation platform for OpenClaw-style AI agents.

Contains clause skills that can be installed to AI agents.

Cascade orchestrates AI agents (Claude Code, Codex, opencode, LLMist) across your workflows in GitHub, Trello, Linear and Jira.

Publish MATLAB functions as MCP tools for AI agents using MATLAB Production Server

Spec-Driven Toolkit to build and deploy Apps/Agents to Azure

Mobile-native agent SDK for building on-device AI agents with sandboxed execution, extensible tools, and x-app / x-agent / x-channel connectivity.

Filesystem-first LLM-Wiki + Obsidian + Hermes Agent memory system. Markdown source of truth, SQLite FTS5 search, secret scanning, tier-based memory. Built for local LLM setups.

Deep Research skill for Claude Code — 3-cycle multi-agent adversarial ensemble methodology

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

基于 Claude Agent Skills 构建的内容创作助手，支持用户需求澄清、深度研究、配图、AI 痕迹优化、多平台风格等功能，一键生成高质量文章

智能 OnCall Agent 是一个基于 AI 的运维自动化助手。采用gin+eino，使用了RAG，ReAct，Plan-Execute-Replan

CortermTerminal2 是一个面向 AI Coding Agent 的远程终端客户端，让你可以在手机上连接服务器，继续操作 Claude Code、Codex、Copilot 等终端任务。

The official implementation of the paper "AgentLAB: Benchmarking LLM Agents against Long-Horizon Attacks"

Mixture-of-Agents-X - utilizing the MoA framework in agentic coding tools.

AI-native publishing system — build books like software. 6-phase pipeline from blank page to EPUB/PDF, with 16 skills, 7 agents, and 8 genre templates. Claude Code plugin.

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.

Cairn — an AI that walks your system and leaves a trail of tests: UI, API, unit, docs.

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.

fak — the Fused Agent Kernel for Claude Code, Codex, Cursor, and MCP: use fewer tokens and do less repeated work with cache-stable model traffic, context compaction + crash resume, local repeat serving, nanosecond tool policy, and local GGUF inference.

A small self-hosted guest lounge for your AI persona — friends chat via one-time links; you only ever see the AI-written visit summary. CC BY 4.0.

Finch: self-improvement orchestrator — mines session history to detect corrections, breakthroughs, and behavioral patterns.

OpenFang 中文本地化版：面向中文用户的开源 AI Agent OS

AI-native software delivery workflow for Claude Code and Codex: design, change, implement, test, review, and archive.

Apache Camel-style integration framework for .NET. 30+ EIP patterns, 27 transports (Kafka, RabbitMQ, MQTT, S3, gRPC, SFTP, SQL...), fluent C# DSL.

Mindlas catches your coding agent drifting before the bad code lands. Real-time tracking and correction for context rot, unverified done claims, patch sprawl, and tool loops in your Claude Code session.

ReAcTree: Hierarchical LLM Agent Trees with Control Flow for Long-Horizon Task Planning (AAMAS 2026)

AKM — universal Agent Knowledge Management. A tool-agnostic Markdown knowledge OS for AI agents: memory / brain / vault over seven layers, with a Learn Back loop. Works with Claude Code, Codex, OpenClaw, and any agent.

Context Diet is a Claude Code agent skill for measuring system-prompt overhead, reviewing tool payload size, and reducing unused definitions through controlled settings.json configuration.

Local semantic code index and MCP server for Claude Code, Codex, and opencode. Includes semantic search, reuse checks, impact analysis, and a VS Code chat and agent supporting local CLIs and API providers—all powered by tree-sitter, SBERT, and ChromaDB.

ash4d.com is a self-hosted private AI platform for k3s Kubernetes, combining local Qwen3 inference, retrieval-augmented generation, agent tooling, and GPU-enabled services in a GitOps-managed environment.

Research Agent for multi-hop reasoning QA — 3rd place🥉 in 2026 Alibaba Cloud Data+AI Global Competition

Open WebUI Pipe: run Claude Code's agent loop inside chats via the Claude Agent SDK

Self-hosted remote terminal for iPhone, iPad & browser — attach tmux and run Claude Code, Codex, or any CLI agent from anywhere. Free for one server.

Audit, clean, and harden a WordPress site end to end, with an AI agent (Claude/Cursor/Codex/MCP or SSH). Detects & removes hidden malware — cloaking, backdoors, database injection — scores your security posture, and hardens safely. Free, open-source (AGPL-3.0).

Autonomous research loop skill for LLM agents — inspired by Karpathy's autoresearch. Define a goal, let the agent iterate.

MUSE-Autoskill - A self-evolving LLM agent that builds, tests, and reuses reusable skills — getting smarter with every task.

This project is a Real-World Multi-Agent AI System built using LangGraph.

Local-first CLI for reproducible biological analysis, with provenance tracked on every step.

A local-Llm-first coding agent. Runs Qwen 2.5 and Qwen 3