Moltbot Skills Index

Moltbot uses a modular Skills system (inspired by MCP - Model Context Protocol) to interact with the world. Skills are tools that the LLM can invoke to perform actions.

Core Skills

These skills are typically enabled by default or highly recommended for all users.

📂 FileSystem

Allows the agent to read, write, move, and delete files within the allowed directories.

• Permissions: Critical. Ensure you only mount necessary directories in Docker.
• Usage: “Create a new file named test.txt”, “Read the config file”.

🌐 Web Browser (Puppeteer)

Gives Moltbot eyes on the web. It can visit URLs, extract text, and even take screenshots.

• Engine: Headless Chrome (via Puppeteer).
• Usage: “Go to google.com and search for…”, “Read the documentation at docs.astro.build”.

💻 Terminal / Command Line

Executes shell commands.

• Risk Level: High.
• Usage: “Run npm install”, “Check git status”.

Community Skills

(Coming Soon) We are compiling a list of community-developed skills for:

• PostgreSQL Database access
• Slack/Discord Integration
• GitHub Issue management

How to Configure Skills

Skills are configured in your config.json or via environment variables.

{
  "skills": {
    "filesystem": {
      "enabled": true,
      "basePath": "./workspace"
    },
    "browser": {
      "enabled": true,
      "headless": true
    }
  }
}