You can deploy this model and start earning money today!

Manojbatra/
my-model

Name: my-model
Availability: InStock

Views

No views yet

MCP Flask Middleware (Docker-ready)

This repo contains a small Flask microservice (MCP) that forwards inference requests to a Hugging Face Inference Endpoint. It's Docker-ready and intended to act as an integration layer for DeepComplianceAgent.

Quick start (local, dev)

Copy .env.example to .env and fill HF_ENDPOINT_URL and HF_API_KEY.
Build & run with docker-compose:

bash

chmod +x entrypoint.sh
docker-compose up --build

Health: GET http://localhost:8080/health
Inference: POST http://localhost:8080/infer with JSON body (same payload you would send to HF).

Deploying to production

Use container registry (Docker Hub, ECR) and run on your cloud provider (ECS, ECS/Fargate, VM, k8s).
Use a secrets manager rather than .env for HF API keys.
Add TLS termination/load balancer in front of the MCP service.

Environment variables

See .env.example.

Security & ops notes

Do not commit .env to source control.
Limit outgoing concurrency to avoid HF overage costs.
Add authentication to /infer (API key, mTLS) if used in production.

Manojbatra/my-model