AI Agent App

What is AI Agent App?

AI Agent App is a collection of sample applications that demonstrate integrating Google Gemini with various AI frameworks, including Model Context Protocol (MCP), RAG (Retrieval-Augmented Generation), Vertex AI, Flutter, and workflow automation tools like n8n. It is intended for developers exploring Gemini-based agent architectures.

How to use AI Agent App?

The README provides a list of sample directories (e.g., /ask_gemini/, /mcp_server/, /rag/). No explicit installation or configuration steps are given; users are expected to explore each sample’s own instructions within its folder.

Key features of AI Agent App

• Gemini chat sample app (/ask_gemini/)
• MCP Server with OpenWeatherMap and BMI Calculator
• MCP Client and CLI Agent (/mcp_client/, /mcp_host/)
• Kama Sutra RAG with FAISS, ChromaDB, and FastAPI
• Flutter app integrated with Vertex AI
• Agent apps using ADK (Agent Development Kit) and n8n

Use cases of AI Agent App

• Build and test a Gemini agent that calls external APIs via MCP
• Implement RAG with different vector databases (FAISS, ChromaDB)
• Create a Flutter mobile frontend powered by Vertex AI
• Automate agent workflows using n8n or custom CLI agents

FAQ from AI Agent App

What samples are included in AI Agent App?

The repository contains samples for a Gemini chat app, an MCP server (OpenWeatherMap and BMI calculator), an MCP client, a CLI agent, multiple RAG apps (FAISS, ChromaDB, FastAPI), a Flutter app with Vertex AI, and agent apps using ADK and n8n.

Where can I find the source code for each sample?

Each sample has its own subdirectory under the repository root, such as /mcp_server/, /rag/, and /ask_gemini/. The README lists all directories.

What transport does the MCP server use?

The MCP server example in the /mcp_server/ sample uses SSE (Server-Sent Events), as indicated by the description “A sample AI Agent Gemini with MCP Server SSE.”

Does the AI Agent App require any external dependencies?

The README does not list dependencies. Users should refer to the individual sample directories for runtime requirements (likely Python, Node.js, or Flutter SDK depending on the sample).