Agent Machines
Open-source building blocks for configuring and operating AI agent workers
Visit agent-machines.dev ↗What is Agent Machines?
Agent Machines provides a framework for assembling persistent AI agents with pluggable runtimes, sandbox environments, tools, and state management. It supports multiple agent implementations (Claude Code, Codex, Hermes, OpenClaw) and compute providers (Daytona, E2B, Sprites, Vercel), letting you configure workers once and swap infrastructure without rebuilding. The system includes runtime adapters, sandbox provisioning, a browser-based terminal for agent interaction, and a dashboard to monitor multiple agents.
Designed for developers building agent systems, it offers source-first components you can inspect, modify, and extend. Features include persistent worker identity and memory across provider changes, credential management, scheduled tasks, tool and MCP integration, and activity logging. The maker indicates it was built with Claude Code.
Momentum
Daily activity over the last 30 days: upvotes, unique visits, new GitHub stars and Hacker News points.
How do we know it's vibe coded?
vibetrend discovered Agent Machines automatically. Evidence: GitHub Kevin-Liu-01/Agent-Machines; repo has .claude, .cursor. View source.
Similar vibe-coded apps
Open-source control plane for orchestrating multiple AI coding agents
Open-source AI job search agent that runs locally in your CLI
IDE to orchestrate multiple coding agents in one workspace
AI-native engineering philosophy and plugin for coding agents
Open-source runtime layer that turns LLMs into working agents
AI framework that injects project specs into coding agents across 20+ platforms