Bee Agent & Model Context Protocol with Backstage

by evanshortiss

7 stars
359 downloads
Not rated
GitHub

About

A sample MCP Server and Agent that uses it to pull data from a Backstage instanace

Details

Author
evanshortiss
GitHub stars
7
Downloads
359
Categories
AI

- MCP server exposing tools for LLMs to use Backstage APIs.
- Sample agent built with the Bee Agent Framework.
- Integration via static token-based HTTP authentication.
- Runs locally on http://localhost:8080.

Configure Backstage with static token-based authentication. Then prepare and run the MCP server by navigating to server/, installing dependencies (npm i), copying .env.template to .env, editing with valid values, and starting with npm run build && npm start. In a separate terminal, run the agent by changing to agent/, copying .env.template, installing dependencies, and executing npm run build && npm start.

Bee Agent & Model Context Protocol with Backstage

This repository demonstrates how to create a Model Context Protocol (MCP)
server that provides tools that LLMs can use to communicate with a Backstage
instance. A sample Agent, built using the Bee Agent Framework,
is included to demonstrate how to interact with the MCP server.

Run the Demo

Configure Backstage

Enable static, token-based, access to Backstage. This allows a client to
interact with the HTTP API using the token in an authorization header.

appConfig:
  backend:
    auth:
      externalAccess:
      - type: static
        options:
        token: 'notasecuretoken'
        subject: mcp

Run the MCP Server

Install dependencies and prepare environment variables:

cd server
npm i
cp .env.template .env

Edit the .env file to with valid values, then start the server:

```bash

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.