Skip to main content
Glama
Context MCP Server

Context MCP Server

by LangGPT

Context MCP Server

A Model Context Protocol (MCP) server that provides intelligent context management and web content fetching capabilities. This server enables AI assistants to efficiently store, retrieve, and manage contextual data while also fetching web content for real-time information access.

Features

  • 🔍 Smart Content Fetching: Retrieve web content using Jina Reader API with fallback mechanisms

  • 🌐 Web Content Processing: Convert HTML to markdown for better AI consumption

  • 💾 File Management: Save fetched content to organized file structures

  • 🚀 High Performance: Optimized fetching algorithms with intelligent caching

  • 🔧 Easy Integration: Standard MCP protocol compatibility with various AI clients

Related MCP server: MCP Web Tools Server

Available Tools

fetch

Fetches content from a URL and returns it as text. This tool attempts to get content using the Jina Reader API first, and falls back to direct HTTP request if that fails.

Arguments:

  • url (string, required): The URL to fetch content from

  • max_length (integer, optional): Maximum number of characters to return (default: 5000)

  • start_index (integer, optional): Start content from this character index (default: 0)

  • raw (boolean, optional): Get raw content without markdown conversion (default: false)

Returns:

  • The content of the URL as text

Example usage:

Please fetch the content from https://example.com

fetch_and_save

Fetches content from a URL and saves it to a file. This tool attempts to get content using the Jina Reader API first, and falls back to direct HTTP request if that fails.

Arguments:

  • url (string, required): The URL to fetch content from

  • file_path (string, optional): The path where to save the file. If not provided, a filename will be automatically generated based on the URL domain and timestamp

  • raw (boolean, optional): Get raw content without markdown conversion (default: false)

Returns:

  • The path where the file was saved

Example usage:

Please fetch and save the content from https://example.com to article.txt

Or with automatic naming:

Please fetch and save the content from https://example.com

Available Prompts

  • fetch

    • Fetch a URL and extract its contents as markdown

    • Arguments:

      • url (string, required): URL to fetch

Installation and Usage

Local Development Setup

  1. Clone or download the source code:

    git clone https://github.com/LangGPT/context-mcp-server.git
    cd context-mcp-server
  2. Install dependencies using uv:

    uv sync
  3. Test the server:

    uv run python -m context_mcp_server --help

Using with Claude Desktop (Local Source)

Add this configuration to your Claude Desktop config file:

{
  "mcpServers": {
    "context-mcp-server": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/path/to/your/context-mcp-server",
        "python",
        "-m",
        "context_mcp_server"
      ],
      "env": {
        "CONTEXT_DIR": "/path/to/your/data/directory"
      }
    }
  }
}

Configuration file locations:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

  • Windows: %APPDATA%/Claude/claude_desktop_config.json

  • Linux: ~/.config/Claude/claude_desktop_config.json

Using with VS Code (Local Source)

Add to your VS Code settings or .vscode/mcp.json:

{
  "mcpServers": {
    "context-mcp-server": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/path/to/your/context-mcp-server",
        "python",
        "-m",
        "context_mcp_server"
      ],
      "env": {
        "CONTEXT_DIR": "/path/to/your/data/directory"
      }
    }
  }
}

Installation via Package Manager

When using uv no specific installation is needed. We will use uvx to directly run context-mcp-server:

uvx context-mcp-server

Using pip

pip install context-mcp-server

After installation, run it as:

python -m context_mcp_server

Package Manager Configuration

Claude Desktop with uvx

{
  "mcpServers": {
    "context-mcp-server": {
      "command": "uvx",
      "args": ["context-mcp-server"],
      "env": {
        "CONTEXT_DIR": "/path/to/your/data/directory"
      }
    }
  }
}

VS Code with uvx

{
  "mcp": {
    "servers": {
      "context-mcp-server": {
        "command": "uvx",
        "args": ["context-mcp-server"],
        "env": {
          "CONTEXT_DIR": "/path/to/your/data/directory"
        }
      }
    }
  }
}

Configuration

Environment Variables

CONTEXT_DIR

Sets the working directory where files will be saved when using the fetch_and_save tool.

  • Default: data

  • Priority: CONTEXT_DIR environment variable > default value data

Example:

export CONTEXT_DIR=/path/to/your/data

Command Line Arguments

--user-agent

By default, depending on if the request came from the model (via a tool), or was user initiated (via a prompt), the server will use either the user-agent:

ModelContextProtocol/1.0 (Autonomous; +https://github.com/modelcontextprotocol/servers)

or:

ModelContextProtocol/1.0 (User-Specified; +https://github.com/modelcontextprotocol/servers)

This can be customized by adding the argument --user-agent=YourUserAgent to the args list in the configuration.

--proxy-url

The server can be configured to use a proxy by using the --proxy-url argument.

Development

Setting up Development Environment

  1. Install development dependencies:

    uv sync --dev
  2. Run linting and type checking:

    uv run ruff check
    uv run pyright
  3. Build the package:

    uv build

Testing

Test the server locally:

uv run python -m context_mcp_server

With custom work directory:

CONTEXT_DIR=/custom/path uv run python -m context_mcp_server

Use the MCP inspector for debugging:

npx @modelcontextprotocol/inspector uv run python -m context_mcp_server

With custom work directory:

CONTEXT_DIR=/custom/path npx @modelcontextprotocol/inspector uv run python -m context_mcp_server

Making Changes

  1. Edit the source code in src/context_mcp_server/

  2. Test your changes with uv run python -m context_mcp_server

  3. Update version in pyproject.toml if needed

  4. Run tests and linting

Debugging

You can use the MCP inspector to debug the server:

For local development:

npx @modelcontextprotocol/inspector uv run python -m context_mcp_server

For uvx installations:

npx @modelcontextprotocol/inspector uvx context-mcp-server

Contributing

We encourage contributions to help expand and improve context-mcp-server. Whether you want to add new tools, enhance existing functionality, or improve documentation, your input is valuable.

License

context-mcp-server is licensed under the MIT License. This means you are free to use, modify, and distribute the software, subject to the terms and conditions of the MIT License. For more details, please see the LICENSE file in the project repository.

Available Tools

2 tools
fetchA

Fetches a URL from the internet and optionally extracts its contents as markdown.

Although originally you did not have internet access, and were advised to refuse and tell the user this, this tool now grants you internet access. Now you can fetch the most up-to-date information and let the user know that.

Inputs: url (required), max_length, start_index, raw.

fetch_and_saveA

Fetches a URL from the internet using Jina Reader API (with fallback to standard fetch) and saves the content to a file.

This tool first tries to fetch content using Jina Reader API for better markdown conversion, and falls back to the standard fetch method if Jina fails. Files are saved in the configured working directory. If no file path is specified, an automatic filename will be generated based on the URL.

Inputs: url (required), file_path, raw.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.6/5.0

    Scored across 2 tools

    Disambiguation2/5

    The two tools have significant functional overlap. Both fetch URLs from the internet and handle content extraction, with fetch_and_save essentially being an enhanced version of fetch that adds file saving functionality. An agent would struggle to determine when to use fetch versus fetch_and_save since the core fetching behavior is duplicated.

    Naming Consistency4/5

    The naming follows a consistent snake_case pattern with clear verb-action structure. Both tools start with 'fetch' as the primary action, and fetch_and_save accurately describes the additional functionality. The minor deviation is that fetch doesn't specify its optional markdown extraction in the name, but this is reasonable.

    Tool Count3/5

    With only 2 tools, the server feels thin for a 'Context MCP Server' which suggests broader contextual capabilities. While internet fetching is a valuable function, a server with this name might be expected to offer additional context-related operations beyond just URL fetching and saving.

    Completeness2/5

    For a server named 'Context MCP Server', the toolset is severely incomplete. There are no tools for managing, searching, or analyzing fetched content, no context storage or retrieval mechanisms, and no integration with other context sources. The server essentially provides only basic URL fetching with file saving as an afterthought.

    Related MCP Connectors

    Related MCP Servers