Project: AutoClaw

Project Overview

AutoClaw is a hyper-lightweight AI agent designed for massive scale automation in headless/containerized environments. It serves as the ideal "runtime" for executing LLM-driven tasks within Docker containers, allowing users to orchestrate thousands of agents simultaneously for complex parallel workflows.

GitHub: https://github.com/tsingliuwin/autoclaw

Core Philosophy

Docker First: Designed to run inside isolated containers (Alpine/Debian).
Massive Scalability: Low resource footprint enables high-concurrency swarms.
Headless & Non-Interactive: Zero GUI dependencies; optimized for CI/CD and Clusters.

Technology Stack

Runtime: Node.js
Language: TypeScript
Framework: Commander.js
UI: Inquirer (interactivity), Chalk (styling), Ora (spinners)
AI: OpenAI SDK

Directory Structure

src/: Source code
- index.ts: CLI entry point and main loop.
- agent.ts: Agent class handling LLM interaction and tool loop.
- tools.ts: Implementation of tools (Shell execution, File I/O).
dist/: Compiled JavaScript files.

Getting Started

Prerequisites

Node.js installed.
OpenAI API Key (or compatible provider like DeepSeek, LocalLLM).

Installation (Development)

Install dependencies:
```
npm install
```
Build the project:
```
npm run build
```

Installation (User)

npm install -g autoclaw

Updating

npm update -g autoclaw

Configuration

AutoClaw uses a hierarchical configuration system.

Priority Order:

CLI Arguments: (-m)
Environment Variables: (.env, System Vars)
Project Config: (./.autoclaw/setting.json)
Global Config: (~/.autoclaw/setting.json)

Setup: Run autoclaw setup to configure the global JSON settings.

Security: Add .autoclaw/ to .gitignore if using project-level config with secrets.

Usage

Run the tool:

npm start

Or use the CLI command if installed globally:

autoclaw

Features

Natural Language Command Execution: "List all markdown files in this folder."
File Management: "Create a new file called test.txt with 'Hello World'."
Safety: All shell commands require user confirmation before execution.
Context Aware: Automatically detects OS and environment.

2.4 KiB Raw Permalink Blame History