MCP CLI Application

What is MCP CLI Application?

MCP CLI Application is a command-line tool that implements the Model Context Protocol (MCP) to extend Amazon Q’s capabilities with additional tools and context. It is built in Go and is configured as an MCP server for Amazon Q.

How to use MCP CLI Application?

Build and run using ./build.sh and ./run.sh [options]. Configure Amazon Q by adding the server definition to ~/.aws/amazonq/mcp.json. Invoke tools directly from the CLI or through Amazon Q by asking for specific tasks.

Key features of MCP CLI Application

• Calculator tool for basic arithmetic operations
• Prompt registry to manage and retrieve stored prompts
• Rule creator to define development standard rules
• Rules processor to check files against rules
• Google Search tool to fetch top search results
• Wikipedia image search and save tools

Use cases of MCP CLI Application

• Extend Amazon Q with custom tools for development workflows
• Perform quick calculations within a chat interface
• Enforce coding standards by creating and applying rules
• Retrieve pre‑defined prompt templates for specialized tasks
• Search the web or fetch Wikipedia images without leaving the IDE

FAQ from MCP CLI Application

How do I configure Amazon Q to use this server?

Add a new entry under mcpServers in ~/.aws/amazonq/mcp.json, specifying the command path to the compiled binary and a tools object for each available tool.

What runtime or dependencies are required?

The application is built with Go. Pre‑built binaries can be produced using ./build.sh; no additional runtime dependencies are mentioned.

Where are prompts stored?

Prompts are stored as files in the ~/.mcp/prompts directory on the local file system.

What tools are available?

The server provides seven tools: calculator, prompt registry, rule creator, rules processor, Google search, Wikipedia image, and Wikipedia image save.

How does Amazon Q use the prompt registry?

Amazon Q can retrieve a prompt template from the registry, fill in variables based on the user’s request, and use the completed prompt to generate more accurate and consistent responses.