Cv MCP

🚀 Frank MCP Personal Resume Server

Frank MCP Personal Resume Server is a microservice built on the FastMCP protocol, mainly used to provide personal resume data and related functions.

🚀 Quick Start

Frank MCP Personal Resume Server is a microservice constructed based on the FastMCP protocol, dedicated to offering personal resume data and relevant functionalities. The server implements multiple tools, including retrieving personal profiles, skill lists, searching resume content, and obtaining media resource links.

✨ Features

Implemented Tools

The following are the main tools implemented by the Frank MCP Personal Resume Server:

• get_profile: Retrieve basic personal information.
• get_skills: Retrieve the list of personal skills.
• search_cv: Search for specific content in the resume.
• get_company_experience: Retrieve work experience at a specified company.
• get_resume_link: Obtain the download link for the personal resume PDF file.
• get_picture_link: Obtain the download link for the personal profile picture file.

Project Structure

The server is built on Node.js and the Bun language, mainly consisting of the following parts:

1. Server: An HTTP service built with Bun, listening on port 3001, and providing two connection methods: standard input-output (stdio) and Server-Sent Events (SSE).
2. Tools: Implements the various tools mentioned above, with each tool corresponding to an independent function or module.
3. Resources: Provides some static resources, including the personal resume PDF file and the profile picture file.

Project Configuration

The root directory of the project contains the following key files:

• package.json: Dependency management file for Node.js and Bun.
• Dockerfile: Configuration file for Docker container construction.
• mcp.json: Configures the information of the MCP server, including connection methods and environment variables.

💻 Usage Examples

Connecting to the MCP Service

Connect via the CLI Client:

# Connect via stdio mode
npx fastmcp dev "bun run start"

# Connect via SSE mode
npx fastmcp connect http://localhost:3001/sse

Configuration in Cursor:

1. Open Cursor and go to the settings interface.
2. Click the "Functions" option on the left sidebar.
3. Scroll to the "MCP Server" section.
4. Click the "Add New MCP Service" button.
5. Fill in the following information:
   • Service Name: frank-cv-mcp
   • For stdio mode:
     • Type: command
     • Command: bun run start
   • For SSE mode:
     • Type: url
     • URL: http://localhost:3001/sse
6. Click "Save".

Configuration Using mcp.json:

{
  "mcpServers": {
    "frank-cv-stdio": {
      "command": "bun",
      "args": ["run", "start"],
      "env": {
        "NODE_ENV": "development"
      }
    },
    "frank-cv-sse": {
      "url": "http://localhost:3001/sse"
    }
  }
}

Request Flow Examples

Obtaining Basic Information:

// Obtain personal profile
const profile = await client.callTool("get_profile", {});

// Obtain skill list
const skills = await client.callTool("get_skills", {});

// Obtain the Markdown format document of the resume
const cvMarkdown = await client.accessResource("cv://frankgoortani");

Searching for Specific Experiences:

// Search for AI-related experiences
const aiExperience = await client.callTool("search_cv", {
  query: "ai"
});

// Obtain work experience at a specified company
const uberExperience = await client.callTool("get_company_experience", {
  company: "Uber"
});

Obtaining Media Links:

// Obtain the download link for the resume PDF
const resumeLink = await client.callTool("get_resume_link", {});

// Obtain the download link for the profile picture
const pictureLink = await client.callTool("get_picture_link", {});

📦 Installation

Docker Usage

This project can be packaged as a Docker image, suitable for Synology NAS or other container orchestration platforms.

Prerequisites

• Docker is installed on your system.

Building the Docker Image:

1. Switch to the project root directory (where the Dockerfile is located).
2. Run the following command to build the image. You can modify the image name if needed:

docker build -t cv-mcp .

Running the Docker Container:

1. After the build is complete, run the container:

docker run -d -p 8080:3001 --name cv-mcp frank_cv_mcp