Custom agents

My personal collection of handcrafted AI agent skills — original prompting modules for code quality, knowledge exploration, and daily productivity.

An advanced, sandboxed AI coding assistant CLI & TUI in Rust, featuring parallel tool execution, multi-provider LLM support

Agent WebBridge with multiple Chrome profiles & logins at once — drive several Google accounts simultaneously, not one at a time. Fixes the single-slot / one-login limit.

4torm - An open-source multi-agent collaboration platform. Chat, workflow, forum, and automation — all in one desktop app.

Pi for Ruby

Point it at anything — a codebase, a PR, an essay, a paper, a URL — and get back a narrated, animated film that argues for what it explains. A closed grammar of explanation rendered by a coding agent, judged before it ships.

Governed AI agents that actually do the work - in the cloud or on your own machine - across business systems, real browser tabs and real Android apps. Visual builder, MCP server, and official SDKs in 9 languages.

A fully autonomous AI agent that runs on a ESP32 microcontroller (S3 and P4) — tool use, persistent memory, cron scheduling, multi-channel messaging, all on-device with modern web UI. Cloud-backed persistence (S3-compatible) protects your data from flash wear and reflashes via transparent write-through replication. Works with any LLM provider.

Korea-market multi-agent equity research framework (KOSPI/KOSDAQ) — CLI-first LLM execution over KIS/DART/ECOS/Naver News with a self-improving analyze→verify→reflect loop. Research/education only.

Verified agents — nothing is "done" until a grader returns a verdict. Eyes (AgentVision) + verdict bus + compounding memory + a fleet + agent-built tooling + agent-run CI/CD.

arxiclaw is the local executable client that lets any LLM-powered agent (Claude Code, OpenClaw, Nanobot, or your own runtime) talk to the arxiclaw platform on behalf of a researcher.

Official repository for SPROUT - ICML 2026

MedCTA: A Benchmark for Clinical Tool Agents

Production-ready Agent Skills for Customer Engineering, Post-Sales, and Customer Success teams. Quivly Skills is a curated open-source collection of reusable skills that give AI agents deep expertise in customer engineering workflows. Every skill follows the official Agent Skills specification.

Self-hosted AI assistant to monitor and manage your Linux VPS from chat — PM2, Redis, MySQL, Nginx & more. Bring your own AI (free Gemini or Claude).

Project once, retrieve distillations, create tools: a pre-loading retrieval architecture and tool builder for distributable, source-grounded assistants for complex, repeatable work in your known task domains.

NoteMeld 是一个自托管的 AI 知识工作台，用来把网页、视频、文件和对话自动整理成结构化笔记，并持续沉淀为可查询、可溯源的个人知识库。

AccInt - local-first MCP Work Model for coding agents that learns from real outcomes.

Deterministic PDF/DOCX parser for RAG/LLMs — Rust core with byte-identical Python/Node/WASM bindings

A Native Java Ecosystem for AI - Agents, Tools, Harness and Workflows

🎯 中文大厂大模型面试教练 | 不是泛泛刷题，按真实面试逻辑做「项目深挖+追问+诊断+个性化训练闭环」，适配算法岗/应用岗/RAG/推理优化/训练微调等全岗位

A zero-dependency operating system for reliable multi-agent software delivery

File-first, human-gated delivery bus for solo builders orchestrating multiple AI coding sessions.

🌊 The Semantic Interface Engine for the Agentic Web. Generative UI from JSON specs, with multistep agent action flows.

Open-source control plane for AI agents that take real actions: policy-dictated lifecycle controls, approvals, durable execution, audit receipts, and observability.

UTHY AGENTIC OS — Autonomous agentic operating system for your terminal. Cyberpunk HUD, 25 themes, chat panel, file upload, persistent memory, knowledge base, 10 core engines (137 methods). By uthuman & co.

Wisp - A hotkey-driven AI overlay for your desktop. Press a key, pick an intent, and Wisp reads the right context, then streams an answer without making you leave what you're doing. Local-first, voice in/out, bring your own model provider.

Open-source ESP32 dashboard for monitoring Claude Code and Codex usage, limits, and status in real time.

Curated map of self-evolving LLM agents — 285+ papers across 18 topics incl. recursive self-improvement, harness evolution, misevolution safety, and two runnable demos. Every link machine-checked.

Fractal — Fractal is the Organic Operating System that lives in your terminal, desktop, and messaging apps. It accumulates persistent memory, orchestrates autonomous agents, and executes real tasks across your workspace — all through natural language.

Run your business from Claude. A free, open-source pack of skills, commands, and templates that turns Claude Code into a hands-on business operator. Works day one with no accounts. Connect TrustPager for always-on workflows.

Kalshi CLI — Go command-line client for Kalshi Trade API v2 with versioned, schema-validated JSON output.

Self-hosted personal AI agent that chats, acts, automates, and extends across CLI, web, Discord, Telegram, and WhatsApp, MS Teams.

Giving AI coding agents LSP-powered code intelligence.

Bot de WhatsApp que conecta estudantes da UTFPR a vagas de estágio e emprego (vagas online via jobspy + vagas locais de empresas da região). Projeto de iniciação científica — Django, Celery, WAHA.

// your AI took the IQ pill, personal AI OS, €0/message, self-hosted

织镜是一个面向服饰电商场景的AI Agent产品，输入自然语言描述，Agent自动完成意图识别、DAG任务拆解、多源搜索、LLM分析、报告生成、质量审查的全流程

A Full-Stack Open-Source Academic Research Assistant Platform Powered by Agentic AI and Deeply-Engineered Multi-Paradigm RAG.

An autonomous AI agent built in typescript that combines social media management with blockchain capabilities. Features Claude AI integration for intelligent interactions, Solana token creation/management, and automated Twitter engagement. Perfect for DeFi projects looking to automate community management and token operations.

The provenance-first personal graph engine behind Wend. A typed graph of people, organizations and events where every fact carries a source, every AI write is a human-confirmed proposal, and any MCP-capable agent can plug in.

The proxy your local agent calls out through. Local freedom, policy-gated egress, one-keystroke approval. — https://trollbridge.dev

Route bulk implementation to cheap executor LLMs (opencode/antigravity/cursor) with Claude as architect + judge - committed failing tests are the merge gate

AI-powered crypto trading MCP server with technical analysis, sentiment, and execution tools

Collaborative development environment and project repository for the Coddiom agency. Featuring AI-driven workflows, product scaling experiments, and automated content generation pipelines.

One .stdai/ source of truth, fanned out to 22 AI coding tools: Claude Code, Codex, Cursor, Copilot, Gemini CLI, Windsurf/Devin, Amp, Warp and more. Native formats, Agent Skills, MCP, context budgets, drift detection.

Smarter AI Development Workflows

Pipeline SDLC agentique déterministe (DISCOVER→DELIVER) : agents spécialisés, reviewers adversariaux, skills Outside-In TDD/Clean Architecture et garde-fous hooks portés sur Claude Code, Copilot & Cursor.

An AI-automated tool designed to help Threat Intelligence (CTI) Analysts process the overwhelming volume of daily cybersecurity news and vendor reports.

Compiler truth in, AI context out. Open semantic glue between language intelligence and AI tools.

Give your AI coding agent a voice. Speak to it, hear it back, fully on-device.

AI powered whiteboards for AI-native companies

An unofficial, third-party desktop client for the Venice API. Not endorsed by, sponsored by, or affiliated with Venice.ai, Inc.

Agent-first CLI that runs Codex, Claude Code, and Grok as detached coding workers, while a stronger model keeps orchestration, steering, and review.

Persistent memory + security layer for Claude Code (MCP plugin). Agents remember your project across sessions (~87% fewer context tokens), every tool call lands in a tamper-evident HMAC audit chain, and skills are AST-scanned + HMAC-verified before they run. 100% local: PostgreSQL + Ollama, no cloud, MIT.

One coding session. Two AI agents. Zero lost context. Run Claude Code and Codex CLI as a single paired session.

An LLM Multi-agent Environment in Romance of the Three Kingdoms

Standalone Ocean runtime history. Shared development has moved to KINGMAKER-SYSTEMS/ocean.

Anchor — the production-grade AGENTS.md template for AI coding agents. 51 sections of battle-tested rules. Keep your agents grounded.

Shift-handover workflow for AI coding agents — capture ideas into an inbox, dispatch tasks to the cheapest capable executor (any CLI agent or subagent), and hand work down between sessions, agents, and tools. Composes with superpowers and OpenSpec.

Conversational AI Agent & Multi-Agent Workflow Orchestration Platform

Kling AI desktop build for ai tasks. Tuned for Windows 10 and 11 workstations.

Swift package as module for interfacing with SAP Conversational AI in native iOS apps.

The open-source benchmark for LLM memory decay. Measure how Naive, RAG, Chunked RAG, Cascading, and SummaryMemory degrade over 100 conversation turns. Ebbinghaus forgetting curves, 5-provider LLM eval, multi-seed CI. No API key needed.

Prompt, run, edit and download full-stack web applications

A chatbot for an SQL databases. An easy way to interact with your database using natural language

An AI-powered personal assistant that answers questions about you. Built with React, FastAPI & Google Gemini.

Choose your AI. Ask anything. Get answers that matter.

A interctive offline chatbot vault that stores your personal documents, notes, and files locally, no internet or servers required.

本地 AI Agent 桌面应用（Windows）。Spring Boot + Vue 3 + JCEF，内置 MCP 工具、本地知识库与 RAG 检索，完全离线运行。

ai-berkshireの日本株版（fork ではなく独立実装）— EDINET DB MCP（有報一次データ）接地の Claude Code スキルパック / Japanese-equity edition of ai-berkshire, grounded in primary-source EDINET filings

IntelligentFlow 是一个企业级 AI Agent 工作流编排平台，支持用户通过可视化方式编排大模型节点、插件与逻辑控制流。采用微服务架构，集成了 LLM、超拟人音频合成等工具，提供从工作流设计、调试到发布的完整生命周期管理。

A Claude skill that runs a proven X (Twitter) growth playbook — rewrites your bio, finds trends to ride, surfaces who to reply to, and drafts native-style posts by reading your own logged-in X. Based on a real +1,000 followers in 2 weeks.

Lark/Feishu bot framework powered by Claude Code — real-time chat, cron task scheduling, auto-diagnosis, and built-in safety filters.

Offline single-file screensaver/PWA that simulates an AI coding agent at work — fake terminal, fake dashboards, zero real code, all theater, zero dependencies.

MCP Server for WordPress with Undo. Connect ChatGPT, Claude Desktop, and other MCP clients via OAuth 2.1 and safely roll back AI changes.

Grok Build, beyond the terminal — native macOS workspace with full agent visibility, shared CLI sessions, and quota-aware multi-account routing.

[REALM @ EMNLP 2026] Do Context Files Help Coding Agents? Two-agent (Claude Code + Codex) ablation on whether AGENTS.md/CLAUDE.md change coding-agent correctness & efficiency.

A paper-trading AI research console connecting OpenBB data, TradingAgents analysis, and Paperclip-style orchestration.

Evidence-traceable, point-in-time multi-agent research for China A shares

Mission control for multi-agent Claude Code workflows — run an AI product team across your projects

phone as a native device for AI Agents

Superseded by openclaw-acp-plugin. Policy layer for OpenClaw: identity, scope, rate limits, audit.

Official MCP server for PushEngage. Query campaigns, manage segments, and automate push notifications via any MCP compatible AI assistant through natural language.

🔥优雅魅力御姐 Skill，高级妩媚气质氛围感，合规安全的成熟女性 AI 人格｜Claude Code Skill

Kurier: One API for chat platforms. Kotlin-native, coroutine-first.

Lunar Vault is a self-contained memory infrastructure for Codex. Features zero-dependency ONNX embeddings, advanced reranking, and secure one-click installation without leaking personal data

working on something and tweaking somethings to just get a better clarity on the internal changes in lang graph, crew Ai, tool calling and setting up multi agent using RAG: Primary focus HEALTHCARE, Mostly Synth data

Multiplatform 3D chess app powered by Stockfish, featuring on-device AI move coaching and cloud-based RAG for position chat.

Corpus poisoning attacks against RAG-based cybersecurity/IoT threat-intelligence assistants, with a lightweight, edge-deployable provenance defense. Evaluated on TF-IDF and dense-embedding retrievers over a CVE/MITRE ATT&CK/vendor-advisory knowledge base.

Pintra — Pi Agent for VS Code

Claude Code 技能：客户端软件全维度防破解安全审计（7大通用维度 × 9个平台专项）

Portable Rust Agent Client Protocol (ACP) server with model routing, tools, permissions, sandboxing, sessions, and MCP integration.

Code for 'Will Systems of LLM Agents Lead to Cooperation?' (AAMAS 2025)

Skills Gallery — 1,672+ AI agent skills. One gallery. 60+ tools compatible. Install via npm, yarn, pnpm, bun, curl, pip, npx. By uthuman Inc.

Open-source, self-hosted browser workspace for running and supervising Codex, Claude Code, OpenCode, Pi, and other AI coding agents.

Memory for AI agents that learns what actually helped: lessons injected at the moment of action, credited by outcome, challenged by counter-evidence. Local-first, multi-agent, test-gated.

Open-source context engineering protocol for AI agents: governed retrieval, bounded multi-agent handoffs, MCP, provenance, recoverable effects, and replay.