@aramassa/mcp-docs-collector

MCP server for collecting documentation from various sources including Git repositories, local files, Confluence, and web pages.

Installation

npm install -g @aramassa/mcp-docs-collector

Or use with npx:

npx -y @aramassa/mcp-docs-collector@latest collect --config mcp_docs.yaml --output ./output

Quick Start

1. Create a configuration file mcp_docs.yaml:

docs:
  - type: local
    path: ./docs
    resources:
      - name: markdown-docs
        includes: ["**/*.md"]

  - type: git
    repoUrl: https://github.com/example/repo.git
    resources:
      - name: api-docs
        includes: ["docs/**/*.md"]

2. Run the collector:

mcp-docs-collector --config mcp_docs.yaml

Features

• Multiple Source Types: Local files, Git repositories, Confluence, and web pages
• Flexible Configuration: YAML-based configuration with environment variable support
• MCP Protocol: Compatible with Model Context Protocol servers
• Automatic Conversion: HTML to Markdown conversion with content extraction
• Cache Support: Built-in caching for web resources
• Frontmatter Support: Automatic YAML frontmatter extraction from Markdown files

Supported Resource Types

| Type | Description | Use Case | |------|-------------|----------| | local | Local filesystem | Project documentation, source code | | git | Git repositories | External repository documentation | | confluence | Confluence Wiki | Organization knowledge base | | web | Web pages & sitemaps | Public API specs, external docs |

Configuration Example

export_envs:
  - GITHUB_TOKEN
  - CONFLUENCE_PAT

docs:
  - type: local
    path: ./docs
    resources:
      - name: project-docs
        includes: ["**/*.md"]
        excludes: ["node_modules/**"]

  - type: git
    repoUrl: [email protected]:example/repo.git
    branch: main
    resources:
      - name: external-docs
        includes: ["docs/**/*.md"]

  - type: confluence
    baseUrl: https://company.atlassian.net/wiki
    apiKey: ${CONFLUENCE_PAT}
    userEmail: ${USER_EMAIL}
    resources:
      - name: wiki-docs
        spaceKey: DEV
        includes: ["**"]

  - type: web
    baseUrl: https://example.com
    resources:
      - name: website-docs
        sitemapUrl: https://example.com/sitemap.xml
        includes: ["docs/**"]

CLI Options

mcp-docs-collector --config <config-file> [options]

| Option | Description | |--------|-------------| | --config <path> | Path to YAML configuration file | | --env-file <path> | Path to environment variables file | | --help | Display help information |

Documentation

For detailed documentation, configuration options, and examples, visit:

• GitHub Repository
• Full Documentation

License

ISC

Author

Aramassa