Code agents

Research, planning, and implementation workflows for Claude Code

Open-source, AI-powered SOC analyst training platform — self-hosted, multi-role, with 13 built-in scenarios

Never hit a rate limit again.

A Claude Code plugin for stashing ideas mid-session. Push to /stash, keep working, pop them back when you're done.

ESG knowledge graph + RAG agent. Turn long-form sustainability disclosures into cited answers, extracted entities, and a reviewable graph of evidence.

RAG Poisoning Lab — Educational AI Security Exercise

🐝 Distill personal semantics into verifiable AI skills and teachable assets.

👌🏻 Pick code into Markdown. Apply Markdown to code. Built for LLMs and quick backups.

curl for x402 and MPP paid APIs. CLI + MCP proxy that auto-pays any endpoint on Base, Solana, and Tempo. Zero crypto code on the buyer side.

Autonomous AI agent for Home Assistant with multi-model routing, persistent memory, and an OpenAI-compatible API

AI mobile automation starter repo for AIVane. Android REPL beta, Python client, skills, examples, and protocol docs for AI agents.

Automated Interpretability and Feature Discovery in Language Models with AI Agents

A simple collection of tools for better plan management by AI agents on OpenCode.

Keep the agent awake, even when the lid closes.

Open-source AI agent: 11 LLM providers (OpenAI, Claude, Ollama, DeepSeek…), 60+ tools, web search, vision, browser & Android automation, MCP plugins, Second Brain memory — CLI · Telegram · WhatsApp · REST API

Local-first MCP coordination hub. One local daemon serves multiple AI assistants across IDEs — shared meetings, tasks, messages, and project registry. Works with Claude, Cursor, VS Code, Zed.

AI browser automation conductor — type a prompt, watch it happen live, get a reusable skill

Residential proxy SDK for Python. 175M+ IPs across 195+ countries. pip install iploop

AI agent skill for searching flights, tours & excursions with real prices and booking links. Works with Claude Code, Cursor, Gemini CLI, Copilot and 30+ AI agents.

Run Claude Code as a headless AI agent over HTTP — hooks + ai + flow. Automate code generation, refactoring, bug triage via REST API using your existing Claude Code subscription.

Install Kodus AI Agents in your own infra.

38 engineering discipline skills for Claude Code & OpenAI Codex — research, quality gates, testing, deployment, gap analysis, and more. Born from real pain, not theory.

7 AKU-architected skills for OpenCode + Obsidian MCP: obsidian-mcp, inbox-triage, connection-review, weekly-synthesis, context-maintenance, vault-health-feedback, note-promotion. Knowledge Activation v4 with validators, EDD evals, and CI. Built on obsidian-mcp-server. Works with OpenCode and Claude Code.

Trust and delegation controls for Paperclip agent teams, powered by AgentVeil Protocol.

Automated comic cataloging tool that identifies issues directly from cover images using a vision-language model, then cross-references results with the Grand Comics Database and the ComicVine API to generate structured, high-confidence collection data with minimal manual entry.

Opinionated first-party agent skills, plugins, subagents, slash commands, and MCP servers for the Litestar framework ecosystem — publishable to Claude Code, Gemini CLI, Codex CLI, Cursor, OpenCode, and VS Code/Copilot from a single repo.

Towards Automating Data Access Permissions in AI Agents

Figma → Code (HTML/CSS/Tailwind/tokens), read by any AI agent via MCP. Free, local, no Dev seat, no account.

Self-hosted Docker setup for Paperclip — the open-source AI company orchestration platform. Built from source, reverse-proxy ready.

Persistent memory and identity for AI agents. Free hosted API.

Local-first AI assistant — 9 specialized agents (code, web, debug, security…), 10M-token vector memory, multimodal (vision + local image generation), two-way voice (dictation + read-aloud), slash commands & codebase RAG, agentic VS Code extension, mobile relay via secure tunnel. 100% local with Ollama. Your data, your control.

Production-ready Claude Code skills for building AI agents with Pydantic AI. Includes dependency injection, tools, validators, streaming, multi-agent orchestration, and evaluation framework patterns.

Open-source CLI scanner for agentic AI components such as skills, MCP servers, system prompts

Claude Code 的恢复源码工程。它包含 CLI 主入口、交互界面、工具系统、命令系统，以及围绕上下文压缩、技能加载、会 话恢复、后台任务和本地服务连接等能力的实现。这份源码当前已经可以完整构建，并能够启动 CLI 构建产物。与此同时，它仍然是一份恢复后的工程，不应被视为官方原始内部仓库；部分能力属于恢复态兼容实现，少数受 fe ature gate 或运行环境影响的路径仍需要按具体场景确认。

Self-hostable runtime for persistent autonomous coding agents — KVM-isolated VMs (or rootless containers), A2A-protocol executor with signed AgentCard discovery, AIWG mission dispatch, web dashboard, virtiofs shared storage. Runs on your hardware; no hosted control plane.

Progressive file disclosure for agentic AI - explore files hierarchically with plugin-based analyzers

gemini cli hooks - adding voice on each hook

WEB APP for Tuya, with AI and customizable functions that increase the performance of your IoT system

Local‑first MCP server for multi‑repository semantic code search with Qdrant and llama. Turns your entire workspace into private context for AI coding assistants like Claude Code, Codex, Cursor, Copilot, Antigravity and Windsurf.

AI skill that converts any content into stunning Apple Keynote-style HTML presentations. Cinematic animations, glassmorphism, ambient effects. Single self-contained .html file, zero dependencies.

NordRelay is a secure remote control bridge for coding agents, connecting Codex, Pi, Hermes, OpenClaw and Claude Code to Telegram, Discord, Slack, WebUI with streaming replies, sessions, files, voice, queues, artifacts and access controls.

AI Legal Defense System with 805,000+ legal documents covering the entire US legal system. Features 31 specialized Slack commands, voice support, session memory, and smart caching. Built with Agent OS + Claude-Flow + Claude Code.

基于 Spring AI Alibaba ReAct 的FishAgent全栈智能体应用。具备分层记忆、三路 RAG 检索、知识库闭环、流式对话与多维限流能力。Java + Python + Vue 3 全栈架构。

AI-powered document intelligence platform for research administration

AI dev teammate for GitHub. Picks up issues, asks questions, ships PRs, answers your review comments, resolves conflicts, merges when green. Self-hosted.

一个可移植的历史人物视角 skill，用于分析战略、矛盾、组织与动员；A portable historical perspective skill for analyzing strategy, contradiction, organization, and mobilization.

38 slash commands that turn Claude Code into a real PDLC workflow — PRD → TDD → implement → review → ship. Hard contracts force AI to persist artifacts, write failing tests first, and run self-checks. No more "looks done" in chat.

📰 OpenClaw Skill: Daily AI/Tech digest from 92 Karpathy-curated blogs. Zero deps, no API keys. Agent-native RSS → Score → Summarize pipeline.

Library of multi-agent teams built on Ejentum cognitive harnesses (adversarial code review, eval trio, side-by-side eval, n8n quickstart, harness integration patterns). Drop-in deployable artifacts for any agent runtime.

A comprehensive 10-project curriculum for developers to learn agentic coding mindsets. Covers rapid UI generation, documentation, refactoring, security patching, and performance optimization using state-of-the-art AI coding assistants.

自然 ZIRAN is an open-source security testing framework for AI agents. It discovers dangerous tool chain compositions via knowledge graph analysis, detects execution-level side effects (not just text output), and runs multi-phase trust exploitation campaigns that model real attacker behaviour.

Docker image to run an Ollama local LLM server. Secure by default, all API requests require a Bearer token (auto-generated on first start). OpenAI-compatible API. Supports first-start model pre-pull, NVIDIA GPU (CUDA) acceleration, and persistent model storage. Multi-arch: amd64, arm64.

FloatChat AI is a conversational AI platform for ARGO Ocean Data Discovery and Visualization, providing accurate and interactive tools to explore oceanographic data. Built with Python, LangChain, Pandas, and machine learning models, it makes complex ocean data accessible and actionable for researchers and enthusiasts.

An operating model for autonomous agent teams. Identity, task contracts, failure memory, trust scoring and audit for teams that want to operate agents as an accountable workforce.

Stoneford — the starter world for WorldLines. 10-agent orchestrator · d20 TRPG · runs standalone in WorldLines or drop-in with Claude Code.

Multi-platform AI assistant skills and workflows. Serious engineering. Also fun.

AI-powered network log analyzer with LLM-assisted root-cause analysis (Junos / EOS / FRR). Local Docker Model Runner or Anthropic Claude. Sanitize-before-LLM. 139 tests. MIT.

🤖 Explore essential AI tools for building, automating, and monetizing projects across content, video, e-commerce, and more.

Open source AI infrastructure layer. Eight agents run on every request: security, caching, memory, hallucination detection, tamper-proof audit trail. One line of code. Your data never leaves your server.

MCP server for BookStack — 56 tools covering the full API + semantic vector search. Rust/tokio/axum, dual transport (SSE + Streamable HTTP), OAuth 2.1, pluggable DB (SQLite/PostgreSQL+pgvector).

Lightweight RAG server for the Model Context Protocol: ingest source code, docs, build a vector index, and expose search/citations to LLMs via MCP tools.

Production-ready RAG chatbot that answers customer service questions from company documents. Built with FastAPI, FAISS, Voyage AI, and Groq — deployed on free-tier cloud infrastructure.

基于 DeepSeek 大语言模型的 AI 编程助手 Agent —— 自然语言驱动的终端编码助手。支持自动读写文件、代码搜索、命令执行、Git 管理，复杂任务智能拆解与多 Agent 并行执行。

A tmux-native workspace for multi-agent AI development across Claude Code, Codex, and Antigravity.

OpenAPI spec, SDKs (TypeScript, Python, Go, Kotlin, Swift, C#), CLI, n8n node, and AI agent skills (MCP-ready) for Pachca API

Persistent AI memory with hybrid search and embedded sync. Open, free, unlimited.

Opus-led Claude Code agent team — strategist plans, Sonnet runners read and fix. Live two-way review-and-fix pipeline. Zero API calls. NOW IN RUST!

🔍 SoSkill - 开源 Skill 搜索与聚合 | Open-source Skill Search & Aggregation

短视频智能体(2个版本：本地部署的和云服务api的)

The loop your AI coding agents follow to ship a change, prove it in production, and improve. A companion file to AGENTS.md and SOUL.md.

Hear a short spoken summary after your coding agent answers. Free to install - no API keys, no subscription, no cloud TTS.

Local Claude/Codex PR review loop using existing CLI subscriptions instead of model API keys

Open Cowork is a desktop, cloud and gateway workspace for OpenCode.

Professional SketchUp Ruby API guidelines for AI agents and developers, compiled from official SketchUp repositories.

Multi-agent social simulation as a Claude Code skill. Zero dependencies — predict anything with /crowdcast simulate.

Autonomous supervisor that uses OpenCode or Claude Code to manage agent-of-empires sessions -- observes, decides, and acts without human intervention

AI visibility intelligence platform (like SEO, but for AI: analyze and optimize how LLMs and AI search engines perceive, cite, and recommend brands. Generative Engine Optimization (GEO).

Agentic Working with LibreOffice

AI coding toolbar for any web app — run npx openmagic, select any element, describe a change, review the diff, and it edits your real source code (verified & self-correcting). Visual element editor + AI UI/UX design review. 14 LLM providers (OpenAI, Claude, Gemini…), BYOK, MIT.

MCP server exposing OpenAI's gpt-image-2 for image generation and iterative editing

MCP server for Claude Code that provides email capabilities through IMAP/SMTP. Read, search, compose, and manage emails from any IMAP provider.

MCP server for rclone.

Local desktop dashboard for Claude Code. Reads your JSONL transcripts and surfaces per-prompt cost, tool heatmaps, subagent attribution, cache analytics, and a rule-based tips engine. Rust + Tauri, fully offline, MIT.

A multimodal AI desktop assistant capable of local OS control, image analysis, and natural voice interaction. Powered by Groq, FastAPI, and React.

Markdown-first, long-term memory infrastructure for AI agents. Hybrid BM25 + semantic search across markdown/code files via MCP.

一个有内在生活的虚拟人格 agent — memory, emotions, subjective relationships and a real daily rhythm

Local-first gateway for personal AI agents with persistent memory across multiple LLM providers

Claude Code plugin: four-layer long-term memory (L0→L1→L2→L3 Persona) with local FTS5 + EmbeddingGemma vector hybrid recall, zero external API

MCP server that wraps the Dagster GraphQL API — monitor and operate runs, assets, schedules, sensors, and backfills from any MCP client

Evidence-based review skill for diff, project, and artifact review across code, readiness, architecture, requirements, and risk.

Local AI diagnostic assistant for Windows PC monitoring. 76 intents, hybrid rule and LLM engine, bilingual PL/EN, context time-windowing, anti-hallucination fallback, and time-travel debugging. Built from real community requests. Part of PC_Workman HCK.

Audit a dbt project for what an AI agent will get wrong if you point it at the data today.

Compress Claude Code agents & skills with semantic verification — 6:1 compression, 100% semantic preservation

AI-first fixed-layout document workspaces. Your AI agent writes proposals, whitepapers, theses, books — A4, PDF, web reader.

Attune-AI is a spec-driven meta-orchestration framework designed to establish a deterministic alignment layer between autonomous LLM agents and a production codebase.

Agentic Software Engineering: A Practical Guide for the AI-Native Engineer

A production-grade conversational AI system showcasing backend architecture, prompt engineering, and real-world deployment readiness.

Self-hosted LLM gateway that turns flat-rate AI subscriptions (Claude Code Max, ChatGPT Plus, Codex, Copilot, M365, Gemini) into one OpenAI/Anthropic-compatible API. Unified subscription wallet, OAuth passthrough on /v1/responses, prompt-injection + PII defense, MCP server, semantic cache, time-travel replay.

MITM proxy for Claude Code with real-time TUI, web UI & MCP server — intercept, inspect, and send messages to the Anthropic API. Headless mode for automation.