Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Demo of a customer service agent (Cymbal Air) using LangGraph, Tools, and RAG to interact with Google Cloud Databases via MCP Toolbox.
| Date | Stars |
|---|---|
| 2026-07-31 | 342 |
| 2026-08-06 | 342 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# 
> [!NOTE]
> This project is for demonstration only and is not an officially supported
> Google product.
>
> If you're a Googler using this demo, please fill up this
> [form](https://forms.gle/dJUdMEbUu7k3TmT4A). If you're interested in using our
> hosted version, please fill up this
> [form](https://forms.gle/3AknwhhWv2pWw46Q8).
## Introduction
This project provides a production-quality reference implementation for building
Agentic applications that use [Agents] and [Retrieval Augmented Generation
(RAG)][rag] to query and interact with data stored in Google Cloud Databases.
This demonstration features Cymbal Air, a fictional airline. The application
showcases a customer service assistant that helps travelers manage flights and
find information about San Francisco International Airport (SFO), Cymbal Air's
hub. The agent can answer questions like:
* *Are there any luxury shops in the terminal?*
* *Where can I get coffee near gate A6?*
* *I need to find a gift for my colleague.*
* *What flights are headed to NYC tomorrow?*
[rag]: https://www.promptingguide.ai/techniques/rag
[Agents]: https://www.promptingguide.ai/agents/introduction
## Table of Contents
<!-- TOC depthfrom:2 -->
- [Understanding the demo](#understanding-the-demo)
- [Retrieval Augmented Generation RAG](#retrieval-augmented-generation-rag)
- [Agent-based Orchestration](#agent-based-orchestration)
- [Architecture](#architecture)
- [Deployment](#deployment)
- [Before you begin](#before-you-begin)
- [One-Time Database & Tool Configuration](#one-time-database--tool-configuration)
- [Launch the Toolbox Server Choose One](#launch-the-toolbox-server-choose-one)
- [Running the Agentic Application](#running-the-agentic-application)
- [Clean Up](#clean-up)
- [Customizing Your Tools](#customizing-your-tools)
<!-- /TOC -->
## Understanding the demo
### Retrieval Augmented Generation (RAG)
One of the best tools for reducing hallucinations is to use Retrieval Augmented
Generation (RAG). RAG is the concept of retrieving some data or information,
augmenting your prompt to the agent, and allowing it to generate more accurate
responses based on the data included in the prompt. This grounds the model’s
response, making it less likely to hallucinate. This technique is also useful
for allowing the agent to access data it didn’t have when it was trained. And
unlike fine-tuning, the information retrieved for RAG does not alter the model
or otherwise leave the context of the request - making it more suitable for use
cases where information privacy and security are important.
Cloud databases provide a managed solution for storing and accessing data in a
scalable and a reliable way. By connecting an agent to a cloud database,
developers can give their applications access to a wider range of information
and reduce the risk of hallucinations.
### Agent-based Orchestration
This application uses an Agent-based orchestration model. Instead of a static
chain of calls, the LLM acts as an intelligent Agent that decides which tools to
use and in what order. It is given a set of available tools, each with a
specific function (e.g., `find_flights`, `list_amenities`). Based on the user's
query, the agent reasons about the best tool to use to find the answer. This
"thought process" allows the agent to handle a wider variety of queries and to
break down complex questions into smaller, manageable steps.
### Architecture
The architecture consists of three main components:
1. **Application** -- The user-facing agentic app that orchestrates the
interaction between the user and the agent.
1. **MCP Toolbox** -- [MCP Toolbox](https://github.com/googleapis/genai-toolbox)
is a middleware server that exposes the database operations as a set of
tools. The LLM agent connects to the Toolbox to execute these tools. This
provides a secure, scalable, and modular way to manage database interactions.
1Excerpt of 8,322 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:6da3398957a63594, topic:agents, topic:langgraph
matched fp:6da3398957a63594, topic:llm
matched fp:6da3398957a63594, topic:rag
matched fp:6da3398957a63594, topic:mcp