enterprise-mcp-template: Instructions file for Claude Code

CLAUDE.md — AI Agent Instructions for Enterprise MCP Template

Purpose: This file tells an AI coding agent (Claude, Copilot, etc.) everything it needs to know to create a new production MCP server from this template. Read this file completely before making any changes.

---

What This Template Is

This is a production-ready template for building MCP (Model Context Protocol) servers. MCP is the standard protocol that lets AI assistants call "tools" (functions) on remote servers. This template implements OAuth 2.0 authentication, HTTP API client with retry logic, token management, and all the deployment infrastructure needed for production.

The template is based on battle-tested patterns from the Luxsant NetSuite MCP project that runs in production on Azure.

---

Core Technology Stack

Technology	Version	Role
Python	3.10+	Runtime
FastMCP	>=3.0.0b2	MCP server framework (handles protocol, OAuth proxy, transports)
httpx	>=0.27.0	Async HTTP client (used for upstream API calls)
Pydantic	>=2.0.0	Data validation models
gunicorn + uvicorn	latest	Production ASGI server
Docker	-	Container deployment

---

How to Create a New MCP Server from This Template

Step-by-Step Process

1. Rename Everything

The template uses placeholder names that MUST be replaced:

FIND                    -> REPLACE WITH (example for NetSuite)
────────────────────────────────────────────────────────────
my_mcp_server           -> netsuite_mcp          (Python package name, snake_case)
my-mcp-server           -> netsuite-mcp          (pip package name, kebab-case)
{{PROJECT_NAME}}        -> NetSuite MCP Enterprise  (display name)
{{AUTHOR}}              -> El Paso Labs           (author/org name)
UPSTREAM_               -> NETSUITE_              (env var prefix)
example.com             -> netsuite.com           (API domain)

Files that need renaming/updating:

• src/my_mcp_server/ → rename directory
• pyproject.toml → update name, description, authors, entry points
• docker-compose.yml → update service name, image name
• Dockerfile → update package name in paths
• startup.sh → update module name
• .env.example → update variable names
• All Python files → update imports