Sentra

PROJET MCP GPT MINIMAL
  • python

0

GitHub Stars

python

Language

7 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

You run a minimal MCP server built with FastAPI that exposes an SSE stream compatible with Developer Mode and two demonstration tools, ping and time. It includes a health endpoint for monitoring and can be run locally or deployed behind a reverse proxy. This guide walks you through usage, installation, configuration, and common operational notes so you can get started quickly and keep things running smoothly.

How to use

Connect your MCP client to the SSE endpoint to subscribe to events and send tool calls. The server exposes a dedicated SSE stream for client interactions and a JSON-RPC interface on the same route, allowing you to request the available demo tools.

Two demo tools are available for testing and demonstrations: ping which responds with a pong message, and time which returns the current UTC timestamp. You can initiate a tool call through the client, targeting one of these two tools, then receive the tool result through the SSE flow.

How to install

Prerequisites you need on your machine:

  • Python 3.12 or newer
  • pip (bundled with Python)
  • git
  • curl or wget for health checks

Set up a local development environment and install dependencies:

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Prepare environment configuration if needed. Copy the example environment file and customize settings such as the service name, SSE path, port, and log level.

cp .env.example .env

Start the MCP server locally so you can begin testing right away.

uvicorn mcp.main:app --host 0.0.0.0 --port 8400 --reload

Verify the health endpoint to ensure the service started correctly.

curl -fsSL http://localhost:8400/health

Configuration and notes

The MCP server is configured via an environment file. Typical settings include the service name, SSE endpoint path, port, and logging level. Use these values to tailor the runtime behavior and observability to your environment.

Health checks are exposed at a REST endpoint and can be used by orchestration tools to verify readiness and liveness. The SSE path is used to establish a streaming connection for tool calls and results.

Docker and container orchestration are supported. Use the provided docker-compose setup to run the API and optionally a reverse proxy for HTTPS. When you enable the reverse proxy profile, HTTPS is handled by the proxy layer.

Operational notes include monitoring resource usage with standard tools and ensuring that only the MCP-related services are exposed to avoid unintended access.

Security and deployment tips

If deploying publicly, place the MCP server behind a reverse proxy with TLS termination and enable strict access controls. Review your firewall rules to allow traffic only on the necessary ports and restrict management interfaces.

Regularly monitor logs and health endpoints to detect anomalies early. Consider adding authentication, rate limiting, and metrics for production readiness.

Troubleshooting

If the server does not come up, check container or process logs for startup errors, verify that the port is not in use, and confirm that environment variables are correctly set.

For health checks, ensure the health endpoint returns a successful status and that the SSE stream endpoint is reachable. If the SSE endpoint is blocked by a proxy, verify proxy routing rules and TLS configuration.

Available tools

ping

Responds with a pong message to verify tool availability and responsiveness.

time

Returns the current UTC timestamp in ISO-8601 format to verify time-related functionality and clock synchronization.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational