CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

This is a FastMCP quickstart project demonstrating the Model Context Protocol (MCP) server implementation using FastMCP 2.0. The repository contains two MCP servers:

echo.py - Simple echo server demonstrating basic FastMCP capabilities
prompt_enhancer.py - Production-ready server for improving prompts for Claude Code

Architecture

FastMCP Server Pattern: The codebase follows FastMCP's decorator-based pattern where a single

FastMCP

instance (

mcp

) serves as the server object, and functionality is registered using decorators:

```
@mcp.tool
```
- Defines callable tools that LLMs can invoke
```
@mcp.resource
```
- Exposes data resources (static or templated URIs)
```
@mcp.prompt
```
- Registers reusable prompt templates

The server in

echo.py

demonstrates all three decorator types with echo functionality.

Prompt Enhancer Server

prompt_enhancer.py

is a comprehensive MCP server designed to improve prompts for Claude Code:

Tools (6 functions):

```
enhance_prompt
```
- Main enhancement function with category support (coding/debugging/refactoring/testing)
```
add_technical_context
```
- Adds language, framework, file paths
```
make_specific
```
- Transforms vague prompts into specific, actionable ones
```
add_best_practices
```
- Adds coding standards and constraints
```
analyze_prompt
```
- Scores and analyzes prompt quality (0-100 scale)
```
compare_prompts
```
- Side-by-side comparison of original vs enhanced

Resources (4 endpoints):

```
prompts://best-practices
```
- Comprehensive guide for writing effective prompts
```
prompts://examples/{category}
```
- Examples for coding, debugging, refactoring, testing
```
prompts://templates/{task_type}
```
- Ready-to-use templates (create_function, fix_bug, etc.)
```
prompts://anti-patterns
```
- Common mistakes to avoid

Prompts (3 templates):

```
enhance-for-claude
```
- Base template for prompt enhancement
```
add-context-template
```
- Template for adding technical context
```
make-actionable-template
```
- Template for making prompts concrete

Development Commands

Running the Server

Local development with MCP Inspector (recommended for testing):

# Echo server
fastmcp dev echo.py

# Prompt Enhancer server
fastmcp dev prompt_enhancer.py

This runs the server via

uv run

subprocess with the MCP Inspector for interactive testing. Note: This command only supports STDIO transport.

Run server directly in current environment:

# Echo server
python echo.py
# or
python3 echo.py

# Prompt Enhancer server
python prompt_enhancer.py

By default, runs with STDIO transport.

Run with HTTP transport (for web-based access):

fastmcp run echo.py --transport http --port 8000
# or in code, modify echo.py to include:
# mcp.run(transport="http", host="127.0.0.1", port=8000)

Run with SSE transport (legacy):

fastmcp run echo.py --transport sse --port 8000

Testing

In-memory testing (fastest, no network/subprocess):

from fastmcp import Client
import asyncio

async def main():
    # Import your server
    from echo import mcp

    # Create in-memory client
    async with Client(mcp) as client:
        # Test ping
        assert await client.ping()

        # List and test tools
        tools = await client.list_tools()
        result = await client.call_tool("echo_tool", {"text": "Hello"})
        print(result)

asyncio.run(main())

Testing with HTTP client:

from fastmcp import Client
import asyncio

async def main():
    async with Client("http://localhost:8000/mcp/") as client:
        assert await client.ping()
        result = await client.call_tool("echo_tool", {"text": "Hello"})
        print(result)

asyncio.run(main())

Key Concepts

Transport Mechanisms: FastMCP supports multiple transports:

STDIO (default) - For local CLI integration
Streamable HTTP (modern, default for HTTP) - For web deployments
SSE (legacy) - Server-Sent Events for web deployments

Server Lifecycle: Always use

if __name__ == "__main__":

block with

mcp.run()

for compatibility with various MCP clients.

Resource URIs: Resources support both static (

echo://static

) and templated URIs (

echo://{text}

) for dynamic content.

Deployment

This repository is designed for FastMCP Cloud deployment:

Create account at http://fastmcp.cloud/signup
Connect GitHub account
Select "Clone our template" for automatic deployment

Using Prompt Enhancer

Example Usage with Claude Code

After installing the Prompt Enhancer server in Claude Code:

fastmcp install claude-code prompt_enhancer.py --server-name "Prompt Enhancer"

You can use it to improve your prompts:

Enhance a vague prompt:
- "Use enhance_prompt tool to improve: 'fix the bug' for category 'debugging'"
Analyze prompt quality:
- "Use analyze_prompt to score: 'Create a function to process users'"
Get best practices:
- "Read the prompts://best-practices resource"
See examples:
- "Show me examples from prompts://examples/coding"
Get templates:
- "Get the template from prompts://templates/create_function"

Common Workflows

Before asking Claude Code to write code:

Use
```
analyze_prompt
```
to check your prompt quality
Use
```
enhance_prompt
```
to get suggestions
Read relevant examples from resources
Use the enhanced prompt with Claude Code

For testing: See

HOW_TO_TEST.md

for detailed testing instructions and examples.

Documentation

FastMCP Documentation: https://gofastmcp.com/
MCP Protocol: https://modelcontextprotocol.io/
FastMCP has MCP server tools available via Claude Code's MCP integration for searching docs
Prompt Enhancer testing guide: See
```
HOW_TO_TEST.md
```

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

This is a FastMCP quickstart project demonstrating the Model Context Protocol (MCP) server implementation using FastMCP 2.0. The repository contains two MCP servers:

echo.py - Simple echo server demonstrating basic FastMCP capabilities
prompt_enhancer.py - Production-ready server for improving prompts for Claude Code

Architecture

FastMCP Server Pattern: The codebase follows FastMCP's decorator-based pattern where a single

FastMCP

instance (

mcp

) serves as the server object, and functionality is registered using decorators:

```
@mcp.tool
```
- Defines callable tools that LLMs can invoke
```
@mcp.resource
```
- Exposes data resources (static or templated URIs)
```
@mcp.prompt
```
- Registers reusable prompt templates

The server in

echo.py

demonstrates all three decorator types with echo functionality.

Prompt Enhancer Server

prompt_enhancer.py

is a comprehensive MCP server designed to improve prompts for Claude Code:

Tools (6 functions):

```
enhance_prompt
```
- Main enhancement function with category support (coding/debugging/refactoring/testing)
```
add_technical_context
```
- Adds language, framework, file paths
```
make_specific
```
- Transforms vague prompts into specific, actionable ones
```
add_best_practices
```
- Adds coding standards and constraints
```
analyze_prompt
```
- Scores and analyzes prompt quality (0-100 scale)
```
compare_prompts
```
- Side-by-side comparison of original vs enhanced

Resources (4 endpoints):

```
prompts://best-practices
```
- Comprehensive guide for writing effective prompts
```
prompts://examples/{category}
```
- Examples for coding, debugging, refactoring, testing
```
prompts://templates/{task_type}
```
- Ready-to-use templates (create_function, fix_bug, etc.)
```
prompts://anti-patterns
```
- Common mistakes to avoid

Prompts (3 templates):

```
enhance-for-claude
```
- Base template for prompt enhancement
```
add-context-template
```
- Template for adding technical context
```
make-actionable-template
```
- Template for making prompts concrete

Development Commands

Running the Server

Local development with MCP Inspector (recommended for testing):

# Echo server
fastmcp dev echo.py

# Prompt Enhancer server
fastmcp dev prompt_enhancer.py

This runs the server via

uv run

subprocess with the MCP Inspector for interactive testing. Note: This command only supports STDIO transport.

Run server directly in current environment:

# Echo server
python echo.py
# or
python3 echo.py

# Prompt Enhancer server
python prompt_enhancer.py

By default, runs with STDIO transport.

Run with HTTP transport (for web-based access):

fastmcp run echo.py --transport http --port 8000
# or in code, modify echo.py to include:
# mcp.run(transport="http", host="127.0.0.1", port=8000)

Run with SSE transport (legacy):

fastmcp run echo.py --transport sse --port 8000

Testing

In-memory testing (fastest, no network/subprocess):

from fastmcp import Client
import asyncio

async def main():
    # Import your server
    from echo import mcp

    # Create in-memory client
    async with Client(mcp) as client:
        # Test ping
        assert await client.ping()

        # List and test tools
        tools = await client.list_tools()
        result = await client.call_tool("echo_tool", {"text": "Hello"})
        print(result)

asyncio.run(main())

Testing with HTTP client:

from fastmcp import Client
import asyncio

async def main():
    async with Client("http://localhost:8000/mcp/") as client:
        assert await client.ping()
        result = await client.call_tool("echo_tool", {"text": "Hello"})
        print(result)

asyncio.run(main())

Key Concepts

Transport Mechanisms: FastMCP supports multiple transports:

STDIO (default) - For local CLI integration
Streamable HTTP (modern, default for HTTP) - For web deployments
SSE (legacy) - Server-Sent Events for web deployments

Server Lifecycle: Always use

if __name__ == "__main__":

block with

mcp.run()

for compatibility with various MCP clients.

Resource URIs: Resources support both static (

echo://static

) and templated URIs (

echo://{text}

) for dynamic content.

Deployment

This repository is designed for FastMCP Cloud deployment:

Create account at http://fastmcp.cloud/signup
Connect GitHub account
Select "Clone our template" for automatic deployment

Using Prompt Enhancer

Example Usage with Claude Code

After installing the Prompt Enhancer server in Claude Code:

fastmcp install claude-code prompt_enhancer.py --server-name "Prompt Enhancer"

You can use it to improve your prompts:

Enhance a vague prompt:
- "Use enhance_prompt tool to improve: 'fix the bug' for category 'debugging'"
Analyze prompt quality:
- "Use analyze_prompt to score: 'Create a function to process users'"
Get best practices:
- "Read the prompts://best-practices resource"
See examples:
- "Show me examples from prompts://examples/coding"
Get templates:
- "Get the template from prompts://templates/create_function"

Common Workflows

Before asking Claude Code to write code:

Use
```
analyze_prompt
```
to check your prompt quality
Use
```
enhance_prompt
```
to get suggestions
Read relevant examples from resources
Use the enhanced prompt with Claude Code

For testing: See

HOW_TO_TEST.md

for detailed testing instructions and examples.

Documentation

FastMCP Documentation: https://gofastmcp.com/
MCP Protocol: https://modelcontextprotocol.io/
FastMCP has MCP server tools available via Claude Code's MCP integration for searching docs
Prompt Enhancer testing guide: See
```
HOW_TO_TEST.md
```

CLAUDE.md

CLAUDE.md

Project Overview

Architecture

Prompt Enhancer Server

Development Commands

Running the Server

Testing

Key Concepts

Deployment

Using Prompt Enhancer

Example Usage with Claude Code

Common Workflows

Documentation

Related Skills

Nano Banana Pro

Markdown Converter

1password

CLAUDE.md

Project Overview

Architecture

Prompt Enhancer Server

Development Commands

Running the Server

Testing

Key Concepts

Deployment

Using Prompt Enhancer

Example Usage with Claude Code

Common Workflows

Documentation