Donobu API

Donobu API Overview 🚀

The Donobu API provides a comprehensive suite of tools for automating web browser interactions. It allows you to create, manage, and execute "flows," which are sequences of actions performed in a browser, driven either by an AI agent or by a predefined script.

The API runs locally on localhost:31000/api/ when you have the Donobu Studio app running.

Key Features

• Flow Management: Create, run, monitor, and delete automation flows
• AI-Powered Automation: Leverage powerful GPT models to drive flows autonomously
• Deterministic Replay: Rerun past flows exactly as they occurred
• Configuration Management: Manage GPT configurations, agent assignments, and environment variables
• Rich Data Access: Retrieve detailed information about flow execution, including screenshots and videos

API Base URL

When running Donobu Studio locally, the API is accessible at:

http://localhost:31000/api/

Main API Endpoints

Flows

• GET /api/flows - List all flows with filtering options
• POST /api/flows - Create and start a new flow
• GET /api/flows/{flowId} - Get specific flow metadata
• DELETE /api/flows/{flowId} - Delete a completed flow
• POST /api/flows/{flowId}/rename - Rename a flow
• GET /api/flows/{flowId}/rerun - Get flow configuration for rerunning
• GET /api/flows/{flowId}/code - Generate Playwright script from flow
• POST /api/flows/project - Generate complete Playwright project from multiple flows

Flow Data

• GET /api/flows/{flowId}/images/{imageId} - Get flow screenshots
• GET /api/flows/{flowId}/video - Get flow video recording
• GET /api/flows/{flowId}/tool-calls - List all tool calls for a flow
• GET /api/flows/{flowId}/tool-calls/{toolCallId} - Get specific tool call details

Configuration

• GET /api/gpt-configs - List all GPT configurations
• GET /api/gpt-configs/{name} - Get specific GPT configuration
• POST /api/gpt-configs/{name} - Create or update GPT configuration
• DELETE /api/gpt-configs/{name} - Delete GPT configuration

Environment Variables

• GET /api/env - List all environment variables
• GET /api/env/{key} - Get specific environment variable
• POST /api/env/{key} - Set environment variable
• DELETE /api/env/{key} - Delete environment variable

Agents

• GET /api/agents - List all agent assignments
• GET /api/agents/{name} - Get agent's GPT configuration
• POST /api/agents/{name} - Assign GPT configuration to agent

Tools & System

• GET /api/tools - List available automation tools
• GET /api/version - Get API version
• GET /api/ping - Health check endpoint

Flow Run Modes

The API supports three different flow execution modes:

AUTONOMOUS

The flow is driven by an AI agent (GPT) that autonomously decides what actions to take based on the overall objective. The AI analyzes the current page state, determines the next appropriate action, and executes it without human intervention.

INSTRUCT

The flow waits for explicit user instructions for each action. The user manually directs what tools to call and when, making this mode suitable for guided walkthroughs or when human oversight is required for each step.

DETERMINISTIC

The flow executes a predetermined sequence of actions without any decision-making. It runs through a fixed list of tool calls and completes when all actions are finished. This mode is commonly used for replaying previously recorded flows.

Example: Creating a Flow

Here's a basic example of creating a new autonomous flow:

curl -X POST http://localhost:31000/api/flows \
  -H "Content-Type: application/json" \
  -d '{
    "name": "My Test Flow",
    "targetWebsite": "https://example.com",
    "overallObjective": "Navigate to the contact page and fill out the form",
    "initialRunMode": "AUTONOMOUS",
    "browser": {
      "using": {
        "type": "device",
        "headless": false
      }
    }
  }'

Getting Started

1. Start Donobu Studio: Make sure your Donobu Studio app is running
2. Configure GPT: Set up at least one GPT configuration for AI-powered flows
3. Test Connection: Use the ping endpoint to verify API connectivity
4. Create Your First Flow: Use the flows endpoint to start automating

For detailed API specifications and examples, refer to the OpenAPI schema available in your Donobu Studio application context.