Cloudscape Docs

A Model Context Protocol (MCP) server that provides semantic search over AWS Cloudscape Design System documentation.

python

1

GitHub Stars

python

Language

6 months ago

First Indexed

2 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

{
  "mcpServers": {
    "praveenc-cloudscape-docs-mcp": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/path/to/cloudscape-docs-mcp",
        "python",
        "server.py"
      ]
    }
  }
}

You can run a local MCP server that provides semantic search over Cloudscape Design System documentation. It exposes two tools for agents: search the documentation index and read full document content, enabling you to build capable AI assistants that quickly locate and extract component information.

How to use

Connect to the Cloudscape Docs MCP Server with your MCP client. You will use two primary actions: search the documentation index to surface the most relevant files and read the full content of a chosen document.

How to install

Prerequisites you need on your machine you must have Python 3.13 or newer.

Install the MCP CLI tools and the server package.

Clone the project repository.

Create and activate a Python virtual environment.

Install dependencies and the server locally using the MCP workflow.

Configuration and usage notes

Place your Cloudscape documentation files inside a docs/ directory. Supported formats include Markdown (.md), plain text (.txt), and TypeScript/React files (.tsx/.ts). The server builds an index by scanning docs/ and creates a vector database for fast, file-based semantic search.

Indexing steps create a vector store under data/lancedb/. You can re-run the indexing command any number of times; note that each run rebuilds the index from scratch to ensure consistency across re-indexes.

Run the server once indexing is complete to start serving MCP requests.

MCP client configuration (stdio)

To run the MCP server from your MCP client, use the standard stdio command with the directory path to the MCP project and the server entry file.

{
  "mcpServers": {
    "cloudscape_docs": {
      "type": "stdio",
      "name": "cloudscape_docs",
      "command": "uv",
      "args": ["run", "--directory", "/path/to/cloudscape-docs-mcp", "python", "server.py"]
    }
  }
}

Notes on usage patterns

Use cloudscape_search_docs to surface the top 5 most relevant files with titles and paths. Use cloudscape_read_doc to fetch the full content of a specific documentation file. These tools enable concise search results followed by access to full material on demand.

Available tools

cloudscape_search_docs

Searches the documentation index and returns the top 5 relevant files with titles and paths.

cloudscape_read_doc

Reads the full content of a specific documentation file as requested by the agent.