base on A collection of production-ready Generative AI Agent templates built for Google Cloud. It accelerates development by providing a holistic, production-ready solution, addressing common challenges (Deployment & Operations, Evaluation, Customization, Observability) in building and deploying GenAI agents. # 🚀 Agent Starter Pack  [](https://youtu.be/jHt-ZVD660g) [](https://googlecloudplatform.github.io/agent-starter-pack/) <a href="https://studio.firebase.google.com/new?template=https%3A%2F%2Fgithub.com%2FGoogleCloudPlatform%2Fagent-starter-pack%2Ftree%2Fmain%2Fsrc%2Fresources%2Fidx"> <picture> <source media="(prefers-color-scheme: dark)" srcset="https://cdn.firebasestudio.dev/btn/try_light_20.svg"> <source media="(prefers-color-scheme: light)" srcset="https://cdn.firebasestudio.dev/btn/try_dark_20.svg"> <img height="20" alt="Try in Firebase Studio" src="https://cdn.firebasestudio.dev/btn/try_blue_20.svg"> </picture> </a> [](https://shell.cloud.google.com/cloudshell/editor?cloudshell_git_repo=https%3A%2F%2Fgithub.com%2Feliasecchig%2Fasp-open-in-cloud-shell&cloudshell_print=open-in-cs)  The `agent-starter-pack` is a Python package that provides a collection of production-ready Generative AI Agent templates built for Google Cloud. <br> It accelerates development by providing a holistic, production-ready solution, addressing common challenges (Deployment & Operations, Evaluation, Customization, Observability) in building and deploying GenAI agents. | ⚡️ Launch | 🧪 Experiment | ✅ Deploy | 🛠️ Customize | |---|---|---|---| | [Pre-built agent templates](./agents/) (ReAct, RAG, multi-agent, Live API). | [Vertex AI evaluation](https://cloud.google.com/vertex-ai/generative-ai/docs/models/evaluation-overview) and an interactive playground. | Production-ready infra with [monitoring, observability](https://googlecloudplatform.github.io/agent-starter-pack/guide/observability), and [CI/CD](https://googlecloudplatform.github.io/agent-starter-pack/guide/deployment) on [Cloud Run](https://cloud.google.com/run) or [Agent Engine](https://cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview). | Extend and customize templates according to your needs. 🆕 Now integrating with [Gemini CLI](https://github.com/google-gemini/gemini-cli) | --- ## ⚡ Get Started in 1 Minute Ready to build your AI agent? Simply run this command: ```bash # Create and activate a Python virtual environment python -m venv .venv && source .venv/bin/activate # Install the agent starter pack pip install --upgrade agent-starter-pack # Create a new agent project agent-starter-pack create my-awesome-agent ``` <details> <summary> ✨ Alternative: Using uv</summary> If you have [`uv`](https://github.com/astral-sh/uv) installed, you can create and set up your project with a single command: ```bash uvx agent-starter-pack create my-awesome-agent ``` This command handles creating the project without needing to pre-install the package into a virtual environment. </details> **That's it!** You now have a fully functional agent project—complete with backend, frontend, and deployment infrastructure—ready for you to explore and customize. ### 🔧 Enhance Existing Agents Already have an agent? Add production-ready deployment and infrastructure by running this command in your project's root folder: ```bash uvx agent-starter-pack enhance ``` See [Installation Guide](https://googlecloudplatform.github.io/agent-starter-pack/guide/installation) for more options, or try with zero setup in [Firebase Studio](https://studio.firebase.google.com/new?template=https%3A%2F%2Fgithub.com%2FGoogleCloudPlatform%2Fagent-starter-pack%2Ftree%2Fmain%2Fsrc%2Fresources%2Fidx) or [Cloud Shell](https://shell.cloud.google.com/cloudshell/editor?cloudshell_git_repo=https%3A%2F%2Fgithub.com%2Feliasecchig%2Fasp-open-in-cloud-shell&cloudshell_print=open-in-cs). --- ## 🤖 Agents | Agent Name | Description | |-----------------------------|-----------------------------------------------------------------------------------------------------------------------------------| | `adk_base` | A base ReAct agent implemented using Google's [Agent Development Kit](https://github.com/google/adk-python) | | `agentic_rag` | A RAG agent for document retrieval and Q&A. Supporting [Vertex AI Search](https://cloud.google.com/generative-ai-app-builder/docs/enterprise-search-introduction) and [Vector Search](https://cloud.google.com/vertex-ai/docs/vector-search/overview). | | `langgraph_base_react` | An agent implementing a base ReAct agent using LangGraph | | `crewai_coding_crew` | A multi-agent system implemented with CrewAI created to support coding activities | | `live_api` | A real-time multimodal RAG agent powered by Gemini, supporting audio/video/text chat with vector DB-backed responses | **More agents are on the way!** We are continuously expanding our [agent library](https://googlecloudplatform.github.io/agent-starter-pack/agents/overview). Have a specific agent type in mind? [Raise an issue as a feature request!](https://github.com/GoogleCloudPlatform/agent-starter-pack/issues/new?labels=enhancement) **🔍 ADK Samples** Looking to explore more ADK examples? Check out the [ADK Samples Repository](https://github.com/google/adk-samples) for additional examples and use cases demonstrating ADK's capabilities. --- ## 🌟 Community Showcase Explore amazing projects built with the Agent Starter Pack! **[View Community Showcase →](https://googlecloudplatform.github.io/agent-starter-pack/guide/community-showcase)** #### Extra Features The `agent-starter-pack` offers two key features to accelerate and simplify the development of your agent: - **🔄 [CI/CD Automation](https://googlecloudplatform.github.io/agent-starter-pack/cli/setup_cicd)** - A single command to set up a complete CI/CD pipeline for all environments, supporting both **Google Cloud Build** and **GitHub Actions**. - **📥 [Data Pipeline for RAG with Terraform/CI-CD](https://googlecloudplatform.github.io/agent-starter-pack/guide/data-ingestion)** - Seamlessly integrate a data pipeline to process embeddings for RAG into your agent system. Supporting [Vertex AI Search](https://cloud.google.com/generative-ai-app-builder/docs/enterprise-search-introduction) and [Vector Search](https://cloud.google.com/vertex-ai/docs/vector-search/overview). - **[Remote Templates](docs/guide/remote-templating.md)**: Create and share your own agent starter packs templates from any Git repository. - **🤖 Gemini CLI Integration** - Use the [Gemini CLI](https://github.com/google-gemini/gemini-cli) and the included `GEMINI.md` context file to ask questions about your template, agent architecture, and the path to production. Get instant guidance and code examples directly in your terminal. ## High-Level Architecture This starter pack covers all aspects of Agent development, from prototyping and evaluation to deployment and monitoring.  --- ## 🔧 Requirements - Python 3.10+ - [Google Cloud SDK](https://cloud.google.com/sdk/docs/install) - [Terraform](https://developer.hashicorp.com/terraform/downloads) (for deployment) - [Make](https://www.gnu.org/software/make/) (for development tasks) ## 📚 Documentation Visit our [documentation site](https://googlecloudplatform.github.io/agent-starter-pack/) for comprehensive guides and references! - [Getting Started Guide](https://googlecloudplatform.github.io/agent-starter-pack/guide/getting-started) - First steps with agent-starter-pack - [Installation Guide](https://googlecloudplatform.github.io/agent-starter-pack/guide/installation) - Setting up your environment - [Deployment Guide](https://googlecloudplatform.github.io/agent-starter-pack/guide/deployment) - Taking your agent to production - [Agent Templates Overview](https://googlecloudplatform.github.io/agent-starter-pack/agents/overview) - Explore available agent patterns - [CLI Reference](https://googlecloudplatform.github.io/agent-starter-pack/cli/) - Command-line tool documentation ### Video Walkthrough: - **[Exploring the Agent Starter Pack](https://www.youtube.com/watch?v=9zqwym-N3lg)**: A comprehensive tutorial demonstrating how to rapidly deploy AI Agents using the Agent Starter Pack, covering architecture, templates, and step-by-step deployment. - **[6-minute introduction](https://www.youtube.com/live/eZ-8UQ_t4YM?feature=shared&t=2791)** (April 2024): Explaining the Agent Starter Pack and demonstrating its key features. Part of the Kaggle GenAI intensive course. - **[120-minute livestream demo](https://www.youtube.com/watch?v=yIRIT_EtALs&t=235s)** (March 6, 2025): Watch us build 3 Agents in under 30 minutes using the `agent-starter-pack`! Looking for more examples and resources for Generative AI on Google Cloud? Check out the [GoogleCloudPlatform/generative-ai](https://github.com/GoogleCloudPlatform/generative-ai) repository for notebooks, code samples, and more! ## Contributing Contributions are welcome! See the [Contributing Guide](CONTRIBUTING.md). ## Feedback We value your input! Your feedback helps us improve this starter pack and make it more useful for the community. ### Getting Help If you encounter any issues or have specific suggestions, please first consider [raising an issue](https://github.com/GoogleCloudPlatform/generative-ai/issues) on our GitHub repository. ### Share Your Experience For other types of feedback, or if you'd like to share a positive experience or success story using this starter pack, we'd love to hear from you! You can reach out to us at <a href="mailto:[email protected]">[email protected]</a>. Thank you for your contributions! ## Disclaimer This repository is for demonstrative purposes only and is not an officially supported Google product. ## Terms of Service The agent-starter-pack templating CLI and the templates in this starter pack leverage Google Cloud APIs. When you use this starter pack, you'll be deploying resources in your own Google Cloud project and will be responsible for those resources. Please review the [Google Cloud Service Terms](https://cloud.google.com/terms/service-terms) for details on the terms of service associated with these APIs. ", Assign "at most 3 tags" to the expected json: {"id":"13909","tags":[]} "only from the tags list I provide: [{"id":77,"name":"3d"},{"id":89,"name":"agent"},{"id":17,"name":"ai"},{"id":54,"name":"algorithm"},{"id":24,"name":"api"},{"id":44,"name":"authentication"},{"id":3,"name":"aws"},{"id":27,"name":"backend"},{"id":60,"name":"benchmark"},{"id":72,"name":"best-practices"},{"id":39,"name":"bitcoin"},{"id":37,"name":"blockchain"},{"id":1,"name":"blog"},{"id":45,"name":"bundler"},{"id":58,"name":"cache"},{"id":21,"name":"chat"},{"id":49,"name":"cicd"},{"id":4,"name":"cli"},{"id":64,"name":"cloud-native"},{"id":48,"name":"cms"},{"id":61,"name":"compiler"},{"id":68,"name":"containerization"},{"id":92,"name":"crm"},{"id":34,"name":"data"},{"id":47,"name":"database"},{"id":8,"name":"declarative-gui "},{"id":9,"name":"deploy-tool"},{"id":53,"name":"desktop-app"},{"id":6,"name":"dev-exp-lib"},{"id":59,"name":"dev-tool"},{"id":13,"name":"ecommerce"},{"id":26,"name":"editor"},{"id":66,"name":"emulator"},{"id":62,"name":"filesystem"},{"id":80,"name":"finance"},{"id":15,"name":"firmware"},{"id":73,"name":"for-fun"},{"id":2,"name":"framework"},{"id":11,"name":"frontend"},{"id":22,"name":"game"},{"id":81,"name":"game-engine "},{"id":23,"name":"graphql"},{"id":84,"name":"gui"},{"id":91,"name":"http"},{"id":5,"name":"http-client"},{"id":51,"name":"iac"},{"id":30,"name":"ide"},{"id":78,"name":"iot"},{"id":40,"name":"json"},{"id":83,"name":"julian"},{"id":38,"name":"k8s"},{"id":31,"name":"language"},{"id":10,"name":"learning-resource"},{"id":33,"name":"lib"},{"id":41,"name":"linter"},{"id":28,"name":"lms"},{"id":16,"name":"logging"},{"id":76,"name":"low-code"},{"id":90,"name":"message-queue"},{"id":42,"name":"mobile-app"},{"id":18,"name":"monitoring"},{"id":36,"name":"networking"},{"id":7,"name":"node-version"},{"id":55,"name":"nosql"},{"id":57,"name":"observability"},{"id":46,"name":"orm"},{"id":52,"name":"os"},{"id":14,"name":"parser"},{"id":74,"name":"react"},{"id":82,"name":"real-time"},{"id":56,"name":"robot"},{"id":65,"name":"runtime"},{"id":32,"name":"sdk"},{"id":71,"name":"search"},{"id":63,"name":"secrets"},{"id":25,"name":"security"},{"id":85,"name":"server"},{"id":86,"name":"serverless"},{"id":70,"name":"storage"},{"id":75,"name":"system-design"},{"id":79,"name":"terminal"},{"id":29,"name":"testing"},{"id":12,"name":"ui"},{"id":50,"name":"ux"},{"id":88,"name":"video"},{"id":20,"name":"web-app"},{"id":35,"name":"web-server"},{"id":43,"name":"webassembly"},{"id":69,"name":"workflow"},{"id":87,"name":"yaml"}]" returns me the "expected json"