Summary Gobii Platform provides production infrastructure for AI agents automating browser interactions. It targets developers needing to scale browser-use agents reliably, offering always-on execution, scheduling, and team collaboration via self-hosted or managed cloud options.

How It Works Gobii transforms browser-use agents into production services by providing essential infrastructure via Docker Compose. It orchestrates agent execution with scheduling, email/API triggers, and secret management. The platform integrates with LLM providers for complex web automation, persistent execution, and structured output capture, leveraging the browser-use library.

Quick Start & Requirements

• Prerequisites: Docker with at least 12 GB RAM allocated to its VM and a few GB of disk space.
• Installation: Clone the repository (git clone https://github.com/gobii-ai/gobii-platform.git), navigate into the directory (cd gobii-platform).
• Start: Run docker compose up --build. Setup wizard at http://localhost:8000 for LLM config (OpenAI, etc.) and API keys. Sign in at http://localhost:8000/.
• Optional Profiles: Use docker compose --profile <name> up for scheduling (beat), email (email), or telemetry (obs).
• Local Dev: docker compose -f docker-compose.dev.yaml up (Postgres, Redis, MinIO).
• Docs: DEVELOPMENT.md for local workflow details.

Highlighted Details

• "Always-on" agent execution with scheduling, email/API triggers, and persistent workflows.
• Team collaboration: shared agents, secret management, access control.
• API endpoints expose browser-use capabilities for integration.
• Monitoring, logging, and structured output capture enhance observability.

Maintenance & Community

• Community support via GitHub issues and Discord.
• Contributions follow ruff/black style.
• Gobii Cloud offers governed releases/enterprise support.

Licensing & Compatibility

• Core source code: MIT License (commercial use permitted).
• Gobii name/logo: Trademarks.
• Proprietary modes/non-MIT components require a commercial agreement with Gobii, Inc.

Limitations & Caveats While the core is MIT-licensed, proprietary features necessitate a commercial agreement, potentially limiting full open-source adoption without additional licensing.