Firebase Genkit MCP Server

Genkit is an open source framework developed by Firebase that helps developers build AI-powered applications using familiar code-centric patterns. It provides unified APIs for generating text, media, structured objects, and tool calls from any generative model, along with vector database support for retrieval-augmented generation. Genkit includes enhanced prompt engineering capabilities, AI workflows, and built-in streaming, making it easier to develop, integrate, and test AI features with observability and evaluations.

Overview

Firebase Genkit is a comprehensive framework for building AI-powered applications with support for Node.js and Go. It provides a unified approach to working with generative AI models, vector databases, and AI workflows, making it easier to develop sophisticated AI features in your applications.

Installation

To get started with Genkit, you can install it globally or as a project dependency:

Global Installation

npm install -g genkit

Project Installation

npm install genkit

For Go applications:

go get github.com/firebase/genkit/go

Key Features

Unified Generation API

Genkit provides a consistent API for generating content from any supported AI model, whether you're generating text, media, structured objects, or tool calls.

Vector Database Support

Easily implement retrieval-augmented generation (RAG) with simple indexing and retrieval APIs that work across different vector database providers.

Enhanced Prompt Engineering

Define rich prompt templates, model configurations, input/output schemas, and tools within a single .prompt file, making your AI interactions more maintainable and testable.

AI Workflows

Organize your AI logic into "Flows" - functions designed for observability, streaming, and easy integration with Genkit devtools. These flows can be deployed as API endpoints.

Built-in Streaming

Stream content from your Genkit API endpoints to create responsive user experiences.

Development Tools

CLI

The Genkit CLI helps you run and evaluate your Genkit functions while collecting telemetry and logs:

# Run your application with Genkit telemetry
genkit start -- <command to run your code>

Developer UI

Genkit includes a local developer UI for testing, debugging, and iterating on your AI application. Key features include:

Run: Execute and experiment with Genkit flows, prompts, and queries in dedicated playgrounds
Inspect: Analyze detailed traces of past executions
Evaluate: Review performance metrics and evaluation results

Plugin Ecosystem

Genkit can be extended with plugins for specific AI models, vector databases, and platform integrations:

Node.js plugins: Available on npm with the keyword "genkit-plugin"
Go plugins: Available on pkg.go.dev

You can also create your own plugins following the plugin authoring guides in the documentation.

Use Cases

Genkit is versatile and can be used to build various AI applications:

Intelligent agents: Create agents that understand user requests and perform tasks autonomously
Data transformation: Convert unstructured data into structured formats
Retrieval-augmented generation: Build apps that provide accurate responses by grounding generation with your own data sources

Documentation

For more detailed information, refer to the official documentation:

Sample Applications

The Genkit repository includes several sample applications to help you get started:

AI barista (simple LLM usage)
Chatbot with JavaScript frontend
Restaurant menu Q&A app
Streaming to an Angular frontend
School assistant system with routing and specialized agents