A

Agent Data Bridge

An MCP server that provides data bridging from Spring Boot interfaces and a lightweight Python sandb...

5 views
0 installs
Updated Feb 3, 2026
Not audited
Tools I Recommend
An MCP server that provides data bridging from Spring Boot interfaces and a lightweight Python sandbox for script execution. It enables agents to fetch data as Markdown or Parquet files and perform automated data analysis within a controlled environment.
  1. Click the “Install Now” button to open the FastMCP connection interface — the environment variable (ENV) values below must be filled into that FastMCP ENV / secrets form before installing.

  2. Open the repository’s .env.example (or this README) to use as a reference for variable names and defaults.

  3. In the FastMCP interface add the following required ENVs (name exactly as shown) and how to obtain each:

    1. REST_HOST — Host address where the FastAPI HTTP API will listen. Default: 127.0.0.1. If running in Docker or you want external access use 0.0.0.0.
    2. REST_PORT — Port for the FastAPI HTTP API. Default: 8000.
    3. MCP_HOST — Host address for the MCP (SSE) server. Default: 0.0.0.0.
    4. MCP_PORT — Port for the MCP (SSE) server. Default: 9000.
    5. APP_ID — Spring Boot OAuth2 client_id used by this service to obtain tokens. If you do not have a special client, you may use the project default: agent.
      • If you control the Spring Boot service: locate the OAuth2 client registration in the Spring Boot configuration (application.yml / application.properties or the service’s admin UI / DB) and use that client_id.
      • If you do not control the Spring Boot service: request the client_id from the API/service owner.
    6. APP_SECRET — Spring Boot OAuth2 client_secret. Default (project example): agent.
      • Obtain it the same way as APP_ID (from Spring Boot config, admin UI, or the API provider). Treat this as a secret.
    7. SANDBOX_DIR — Directory path used for sandbox file storage. Default: sandbox_storage (create this folder on the host or bind-mount into the container). Ensure the process has read/write permissions.
    8. (Optional) SPRING_BOOT_BASE_URL and SPRING_BOOT_API_PATH — not required by current code paths (the service uses the host value passed to fetch endpoints), set only if your deployment or tooling requires them.
  4. Mark sensitive values (APP_SECRET and any API keys) as “secret”/encrypted in the FastMCP interface so they are not shown in plain text.

  5. Save the ENV configuration in FastMCP and then click the Install/Provision action in FastMCP to apply the ENVs and start the service.

  6. Verify the service is up:

    1. Call the health endpoint: GET http://:/health — expect {"status":"ok"}.
    2. (Optional) Test a fetch: POST http://:/api/fetch with a small payload (e.g., host pointed at your Spring Boot test instance) to confirm APP_ID/APP_SECRET are valid and sandbox writing works.
  7. If APP_ID/APP_SECRET authentication fails:

    1. Re-check the client_id/client_secret from the Spring Boot service configuration (application.yml / clients DB / admin).
    2. If you don’t control the Spring Boot server, contact the API owner to create or provide an OAuth2 client credential pair and return both values to use in FastMCP.
  8. If running via Docker / docker-compose, ensure SANDBOX_DIR is volume-mounted to the host (e.g., ./sandbox_storage:/app/sandbox_storage) so generated parquet files persist and the sandbox has correct permissions.

  9. After successful verification, keep a secure copy of any secrets outside FastMCP if required by your operational procedures, or rotate credentials per your security policy.

Quick Start

View on GitHub

More for API Development

View All →

More for Data & Analytics

View All →

Similar MCP Servers

Report Issue

Thank you! Your issue report has been submitted successfully.