This project leverages tools within the Google AI agent ecosystem. Below is a comparison of key tools—Agent Development Kit (ADK), Agent Starter Pack, and Agentspace—to help you understand their roles and how they complement each other:
| Feature | Agent Development Kit (ADK) | Agent Starter Pack | Agentspace |
|---|---|---|---|
| Focus | Code-first toolkit for building, evaluating, and deploying custom AI agents with Python. | Production-ready templates for rapid Generative AI agent development. | Enterprise platform for managing and scaling AI agents. |
| Agent Creation | Modular, multi-agent systems with custom tools (e.g., Google Search, A2A protocol). | Pre-built templates (e.g., ReAct, RAG) with customization options. | No-code/low-code via Agent Designer; Agent Gallery for pre-built agents. |
| Deployment | Containerized on Cloud Run or Vertex AI Agent Engine. | Automated CI/CD with observability tools. | Scalable within a secure enterprise platform. |
| Integrations | Gemini, custom APIs, multi-language support. | Vertex AI, Gemini CLI, LangGraph. | Salesforce, Microsoft Copilot, Google Workspace, 1,000+ partners. |
| Access | Open-source (Apache 2.0), free via GitHub. | Open-source (Apache 2.0), free via GitHub. | Enterprise subscription ($25 USD/seat/month after 30-day trial). |
| Best For | Developers building complex, custom multi-agent systems from scratch. | Teams wanting to quickly prototype and deploy production-ready agents using proven templates. | Enterprises scaling agent adoption organization-wide with governance and management. |