AutoClaw is a hyper-lightweight, engineering-first AI agent framework designed for headless, massive-scale concurrency within Docker containers. It addresses the instability and scalability limitations of traditional vision-based AI agents by relying on precise command-driven execution via system APIs and shell commands. This makes it ideal for developers and organizations requiring robust, deterministic automation in complex, containerized environments like CI/CD pipelines or large-scale orchestration platforms.

How It Works

AutoClaw operates on a foundation of precise, command-driven execution, eschewing unstable visual recognition for direct API and shell command interactions. Its Docker-native design ensures a minimal footprint, making it suitable for resource-constrained environments. The framework is stateless, facilitating seamless orchestration across thousands of instances via Kubernetes, Docker Swarm, or simpler scripting methods, enabling true automation swarms with predictable outcomes.

Quick Start & Requirements

• Installation: User: npm install -g autoclaw. Development: git clone ... && npm install && npm run build.
• Prerequisites: Node.js, npm. An OpenAI API key (or compatible LLM endpoint URL and key). Optional: Tavily API key, SMTP credentials, notification webhook URLs.
• Setup: Run the interactive setup wizard: autoclaw setup.
• Usage: Interactive mode: autoclaw. Headless (one-shot): autoclaw "command" --no-interactive. Auto-confirm: autoclaw "command" -y.
• Links: GitHub Repository: https://github.com/tsingliuwin/autoclaw

Highlighted Details

• Headless Execution: Operates purely via terminal commands, requiring no browsers or GUIs.
• Non-Interactive & Auto-Confirm Modes: Supports zero-touch automation (--no-interactive) and automatic approval of tool executions (-y), with the latter noted as dangerous.
• Extensible Integrations: Built-in support for Web Search (Tavily), Email (SMTP), and Notification Webhooks (Feishu, DingTalk, WeCom).
• Context Aware: Detects container environments and provides accurate system time for temporal context.

Maintenance & Community

Contributions are welcomed via pull requests. The project is hosted on GitHub, serving as the primary point for community interaction and development tracking. No specific community channels (e.g., Discord, Slack) or notable contributors/sponsors are detailed in the provided text.

Licensing & Compatibility

• License: MIT.
• Compatibility: The MIT license is permissive, generally allowing for commercial use and integration within closed-source projects without significant restrictions.

Limitations & Caveats

When running within Docker containers (especially Alpine or Debian Slim), Chinese characters and emojis may render as squares ("tofu") due to missing fonts; CJK and Emoji fonts must be manually installed. Storing sensitive API keys or secrets in the local .autoclaw/setting.json configuration file requires careful management (e.g., adding .autoclaw/ to .gitignore) to prevent accidental exposure.