Skip to content

Documentation

Raven documentation

Install and use Raven, connect it to agents and messaging, or build your own extensions.

Start with installation, then choose the path that matches your goal. Guide covers daily work and deployment. Reference defines protocols and runtime behaviour. Developer Guide explains how to build, debug, and extend Raven.

Start here

  • Install Raven

    Install Raven on your platform, configure a model provider, and start using the WebUI.

    Quick Start

  • Self-host Raven

    Deploy Raven on your own machine and configure its services.

    Self-Hosting

  • Open the WebUI

    Open the browser interface and manage the background service.

    Launch WebUI

Use Raven

  • Use Raven

    Work through daily tasks with sub-agents, browser collaboration, knowledge bases, Playbooks, proactive follow-ups, and Oncall.

    Open the guide

  • Connect and extend Raven

    Connect messaging platforms and agents, then add skills, memory, MCP, plugins, and other extensions.

    Explore integrations

  • Deploy and run Raven

    Choose self-hosting or Docker, launch the WebUI, and set boundaries for local tool execution.

    Open deployment guide

Explore the documentation

  • Architecture and Runtime Contracts

    Understand Raven's architecture, protocols, orchestration, permissions, commands, and repository structure.

    Browse the reference

  • Build and Extend Raven

    Build and extend Raven with agents, plugins, adapters, trajectory replay, Evolver, and tracing.

    Open the developer guide

Use the site search or Command Reference to find a feature's setup instructions, examples, and troubleshooting steps.

Raven is developed in the open by EverMind.