PatternFly MCP Server

A Model Context Protocol (MCP) server that provides access to PatternFly React development rules and documentation, built with Node.js and TypeScript.

What is MCP?

The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely access external data sources and tools. This server provides a standardized way to expose PatternFly documentation and development rules to MCP-compatible clients.

Features

• TypeScript: Full type safety and modern JavaScript features
• PatternFly Documentation Access: Browse, search, and retrieve PatternFly development rules
• Comprehensive Rule Coverage: Access setup, guidelines, components, charts, chatbot, and troubleshooting documentation
• Smart Search: Find specific rules and patterns across all documentation
• Error Handling: Robust error handling with proper MCP error codes
• Modern Node.js: Uses ES modules and the latest Node.js features

Prerequisites

• Node.js 20.0.0 or higher
• npm (or another Node package manager)

Installation

Local development

1. Install dependencies:

npm install

2. Build the project:

npm run build

3. Run in watch/dev mode (TypeScript via tsx):

npm run start:dev

Use via npx (after publishing)

npx @cdcabrera/pf-mcp

Or install locally in a project and run:

npm install @cdcabrera/pf-mcp
npx @cdcabrera/pf-mcp

Scripts

These are the most relevant NPM scripts from package.json:

• build: Build the TypeScript project (cleans dist, type-checks, bundles)
• build:clean: Remove dist
• build:watch: Build in watch mode
• start: Run the built server (node dist/index.js)
• start:dev: Run with tsx in watch mode (development)
• test: Run linting, type-check, and unit tests in src/
• test:dev: Jest watch mode for unit tests
• test:integration: Build and run integration tests in tests/
• test:integration-dev: Watch mode for integration tests
• test:lint: Run ESLint (code quality checks)
• test:lint-fix: Run ESLint with auto-fix
• test:types: TypeScript type-check only (no emit)

Usage

The MCP server communicates over stdio and provides access to PatternFly documentation through the following tools. Both tools accept an argument named urlList which must be an array of strings. Each string is either:

• An external URL (e.g., a raw GitHub URL to a .md file), or
• A local file path (e.g., documentation/.../README.md). When running with the --docs-host flag, these paths are resolved under the llms-files directory instead.

Returned content format:

• For each entry in urlList, the server loads its content, prefixes it with a header like: # Documentation from <resolved-path-or-url> and joins multiple entries using a separator: \n\n---\n\n.
• If an entry fails to load, an inline error message is included for that entry.

Tool: usePatternFlyDocs

Use this to fetch high-level index content (for example, a local README.md that contains relevant links, or llms.txt files in docs-host mode). From that content, you can select specific URLs to pass to fetchDocs.

Parameters:

• urlList: string[] (required)

Response (tools/call):

• content[0].type = "text"
• content[0].text = concatenated documentation content (one or more sources)

Tool: fetchDocs

Use this to fetch one or more specific documentation pages (e.g., concrete design guidelines or accessibility pages) after you’ve identified them via usePatternFlyDocs.

Parameters:

• urlList: string[] (required)

Response (tools/call):

• content[0].type = "text"
• content[0].text = concatenated documentation content (one or more sources)

Docs-host mode (local llms.txt mode)

If you run the server with --docs-host, local paths you pass in urlList are resolved relative to the llms-files folder at the repository root. This is useful when you have pre-curated llms.txt files locally.

Example:

npx @cdcabrera/pf-mcp --docs-host

Then, passing a local path such as react-core/6.0.0/llms.txt in urlList will load from llms-files/react-core/6.0.0/llms.txt.

MCP client configuration examples

Most MCP clients use a JSON configuration that tells the client how to start this server. The server itself does not read that JSON; it only reads CLI flags and environment variables. Below are examples you can adapt to your MCP client.

Minimal client config (npx)

{
  "mcpServers": {
    "patternfly-docs": {
      "command": "npx",
      "args": ["-y", "@cdcabrera/pf-mcp@latest"],
      "description": "PatternFly React development rules and documentation"
    }
  }
}

Docs-host mode

{
  "mcpServers": {
    "patternfly-docs": {
      "command": "npx",
      "args": ["-y", "@cdcabrera/pf-mcp@latest", "--docs-host"],
      "description": "PatternFly docs (docs-host mode)"
    }
  }
}

Local development (after build)

{
  "mcpServers": {
    "patternfly-docs": {
      "command": "node",
      "args": ["dist/index.js"],
      "cwd": "/path/to/patternfly-mcp",
      "description": "PatternFly docs (local build)"
    }
  }
}

Inspector-CLI examples (tools/call)

Note: The parameter name is urlList and it must be a JSON array of strings.

usePatternFlyDocs (example with a local README):

npx @modelcontextprotocol/inspector-cli \
  --config ./mcp-config.json \
  --server patternfly-docs \
  --cli \
  --method tools/call \
  --tool-name usePatternFlyDocs \
  --tool-arg urlList='["documentation/guidelines/README.md"]'

fetchDocs (example with external URLs):

npx @modelcontextprotocol/inspector-cli \
  --config ./mcp-config.json \
  --server patternfly-docs \
  --cli \
  --method tools/call \
  --tool-name fetchDocs \
  --tool-arg urlList='[
    "https://raw.githubusercontent.com/patternfly/patternfly-org/refs/heads/main/packages/documentation-site/patternfly-docs/content/design-guidelines/components/about-modal/about-modal.md",
    "https://raw.githubusercontent.com/patternfly/patternfly-org/refs/heads/main/packages/documentation-site/patternfly-docs/content/accessibility/components/about-modal/about-modal.md"
  ]'

Environment variables

• DOC_MCP_FETCH_TIMEOUT_MS: Milliseconds to wait before aborting an HTTP fetch (default: 15000)
• DOC_MCP_CLEAR_COOLDOWN_MS: Default cooldown value used in internal cache configuration. The current public API does not expose a clearCache tool.

Programmatic usage (advanced)

The runServer function is exported via the package exports map. You can import it from the dedicated subpath or the package root.

Example (ESM):

// Prefer the public export subpath
import { runServer } from '@jephilli-patternfly-docs/mcp/server';

// Or from the package root (index.ts re-exports it)
import { runServer } from '@jephilli-patternfly-docs/mcp';

// Starts the MCP server and listens on stdio
await runServer();
// Server runs until interrupted (Ctrl+C)

Returned content details

For each provided path or URL, the server returns a section:

• Header: # Documentation from <resolved-path-or-url>
• Body: the raw file content fetched from disk or network
• Sections are concatenated with \n\n---\n\n

This makes it easier to see where each chunk of content came from when multiple inputs are provided.

Publishing

To make this package available via npx, you need to publish it to npm:

1. Ensure you have an npm account and are logged in:

npm login

2. Update the version in package.json if needed:

npm version patch  # or minor/major

3. Publish to npm:

npm publish

After publishing, users can run your MCP server with:

npx @cdcabrera/pf-mcp

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

License

MIT License - see LICENSE file for details.

Resources

• Model Context Protocol Documentation
• MCP SDK Documentation
• Node.js Documentation
• TypeScript Documentation