Skip to content

Configuration

Dodo is designed to be flexible. You can configure it to use almost any LLM provider and customize safety settings.

The easiest way to configure Dodo is using the built-in interactive wizard.

  1. Start Dodo: 
    dodo
  2. Type /conf in the input bar (or select Configure from the main menu starting screen).

The wizard will guide you through: - Selecting a Provider: OpenAI, Anthropic, Gemini, or Local. - Entering API Keys: Securely stored on your device. - Sandbox Settings: Enabling Docker for safe execution.

Configuration is saved to ~/.dodo/config.yaml.

Manual Configuration

You can also create or edit the config file manually at ~/.dodo/config.yaml.

LLM Providers

llm:
  provider: openai
  api_key: sk-your-api-key
  model: gpt-4o

llm:
  provider: anthropic
  api_key: sk-ant-your-key
  model: claude-3-5-sonnet-20240620

You can use any OpenAI-compatible endpoint (like Ollama or LM Studio).

llm:
  provider: openai
  base_url: http://localhost:11434/v1
  api_key: ollama  # Value doesn't matter for Ollama, but must be present
  model: llama3

Safety & Sandboxing

Dodo strongly recommends using Docker to sandbox the agent's execution. This prevents the agent from accidentally (or intentionally) modifying files outside your project or installing unwanted software.

To enable sandboxing:

  1. Ensure Docker Desktop is installed and running.
  2. Set sandbox_mode: auto in your config.
sandbox:
  mode: auto  # Use "docker" to force fail if docker is missing

Without Sandbox

If you disable sandboxing (mode: host), the agent runs commands properly on your machine shell. Only use this with trusted models.