Self-hosted multi-agent system that autonomously plans and runs penetration tests using tools like nmap, Metasploit and sqlmap in Docker.
An agent meta-harness that adds multi-agent swarms, shared memory, and coordination around Claude Code and Codex via npx and MCP.
A multi-platform proxy client built on the mihomo (Clash.Meta) core, for Android, Windows, macOS and Linux.
Open-source Node.js and React app that orchestrates teams of AI agents with org charts, budgets, goals, and a task-manager interface.
Cross-platform desktop GUI client for the Clash Meta (mihomo) proxy core, built on Tauri 2 for Windows, macOS and Linux.
Open-source LLM knowledge platform that turns documents into a searchable RAG system, a multi-step reasoning agent, and a maintained wiki.
Claude Code plugin that runs a GitHub Projects board through build, QA, and review skills, merging PRs when checks pass.
OpenShell is a sandboxed runtime that enforces file, system call, and network policies on autonomous AI agents.
Autonomous API and web-app penetration testing tool that runs a swarm of AI agents to chain exploits and prove vulnerabilities.
Open-source system where LLMs autonomously generate research ideas, run experiments, and write full scientific papers.
Flutter GUI proxy client for the mihomo (Clash.Meta) core, available on iOS, macOS, Android, Windows and Linux.
Orchestrates isolated, autonomous coding-agent runs that pick up tasks from a project board such as Linear and deliver pull requests.
Declarative Kubernetes-style orchestrator that runs sandboxed autonomous agent tasks on Agent Substrate, with workspaces and model specs.
Shell script tool that deploys and manages mihomo and sing-box proxy kernels as a client on Linux and router devices.