tamim-mcp-server

tamim-mcp-server

Converts images to PDF with support for local files, URLs, and base64 input, compatible with any AI agent via HTTP or stdio transport.

Category
访问服务器

README

tamim-mcp-server

A local MCP (Model Context Protocol) server that converts images to PDF. Supports any AI agent via two transport modes:

  • Streamable HTTP — for Cursor, Claude Code, and any HTTP-capable MCP client
  • stdio — for Claude Desktop, VS Code MCP, Cline, Roo Code, Windsurf, Continue, and any stdio-spawning agent

What's inside

Tool Description
convert_images_to_pdf Convert local files, URLs, or base64 data into a single PDF
get_server_time Sanity-check tool to confirm the connection is working

File structure

tamim-mcp-server/
├── src/
│   ├── server.ts            # Fastify HTTP transport (Streamable HTTP)
│   ├── stdio-server.ts      # stdio transport
│   ├── mcp.ts               # createMcpServer() factory
│   ├── config.ts            # env vars (port, token)
│   └── tools/
│       ├── index.ts         # registerTools() - wires every tool in
│       ├── time.ts          # get_server_time
│       └── imagesToPdf.ts   # convert_images_to_pdf
├── .env.example
├── .gitignore
├── package.json
├── tsconfig.json
└── README.md

1. Install

npm install
cp .env.example .env
# open .env and set MCP_TOKEN (openssl rand -hex 24), or leave it blank for no auth

2. Run it

npm run dev      # tsx watch, auto-reloads on save
# or, for a built version:
npm run build && npm start

You should see:

MCP server ready on http://127.0.0.1:8787/mcp

3. Test it before touching any client

curl -i -X POST http://127.0.0.1:8787/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"curl","version":"1"}}}'

A 200 OK with a mcp-session-id header back means the server is healthy. Or use the official MCP Inspector for a UI instead of raw curl:

npx @modelcontextprotocol/inspector
# point it at http://127.0.0.1:8787/mcp

4. Connect any AI agent

HTTP agents (server must already be running)

Cursor

.cursor/mcp.json (project root) or ~/.cursor/mcp.json (global):

{
  "mcpServers": {
    "tamim-local-mcp": {
      "url": "http://127.0.0.1:8787/mcp",
      "headers": { "Authorization": "Bearer YOUR_TOKEN_HERE" }
    }
  }
}

Omit the headers block entirely if you left MCP_TOKEN blank.

Claude Code

claude mcp add --transport http tamim-local-mcp http://127.0.0.1:8787/mcp \
  --header "Authorization: Bearer YOUR_TOKEN_HERE"

stdio agents (auto-spawned, no manual server start needed)

Claude Desktop

claude_desktop_config.json:

{
  "mcpServers": {
    "tamim-local-mcp": {
      "command": "npx",
      "args": ["-y", "tamim-mcp-stdio"]
    }
  }
}

If you installed locally instead of publishing to npm:

{
  "mcpServers": {
    "tamim-local-mcp": {
      "command": "node",
      "args": ["dist/stdio-server.js"],
      "cwd": "/absolute/path/to/tamim-mcp-server"
    }
  }
}

VS Code / GitHub Copilot

.vscode/mcp.json (project root):

{
  "servers": {
    "tamim-local-mcp": {
      "command": "node",
      "args": ["dist/stdio-server.js"],
      "cwd": "${workspaceFolder}"
    }
  }
}

Cline / Roo Code

~/.cline/mcp.json or ~/.roo/mcp.json:

{
  "mcpServers": {
    "tamim-local-mcp": {
      "command": "node",
      "args": ["dist/stdio-server.js"],
      "cwd": "/absolute/path/to/tamim-mcp-server"
    }
  }
}

Windsurf

.windsurf/mcp.json:

{
  "mcpServers": {
    "tamim-local-mcp": {
      "command": "node",
      "args": ["dist/stdio-server.js"],
      "cwd": "/absolute/path/to/tamim-mcp-server"
    }
  }
}

Continue

.continue/config.yaml:

mcpServers:
  - name: tamim-local-mcp
    command: node
    args:
      - dist/stdio-server.js
    cwd: /absolute/path/to/tamim-mcp-server

Generic stdio config (works with any MCP client)

{
  "command": "node",
  "args": ["dist/stdio-server.js"],
  "cwd": "/absolute/path/to/tamim-mcp-server"
}

5. Tool usage

convert_images_to_pdf

Accepts images from three sources (at least one required):

// Local file paths
{
  "imagePaths": ["/path/to/photo1.jpg", "/path/to/photo2.png"],
  "outputPath": "/path/to/output.pdf"
}

// URLs
{
  "urls": ["https://example.com/image1.jpg", "https://example.com/image2.png"],
  "outputPath": "/path/to/output.pdf"
}

// Base64 data
{
  "base64Images": ["/9j/4AAQSkZJRg...", "iVBORw0KGgo..."],
  "outputPath": "/path/to/output.pdf"
}

// Mix of all three
{
  "imagePaths": ["/path/local.jpg"],
  "urls": ["https://example.com/photo.png"],
  "base64Images": ["iVBORw0KGgo..."],
  "outputPath": "/path/to/output.pdf"
}

Options:

  • pageSize: "fit" (default) — each page matches its image, or "a4" — centers on A4
  • marginPt: margin in points when using A4 (default: 24)

get_server_time

No parameters. Returns the current server time in ISO format.


Keeping it running (HTTP mode)

HTTP-based MCP servers don't get auto-started by the client the way stdio ones do - the server has to already be running when HTTP clients try to connect. Options, easiest first:

  • Leave npm run dev in a spare terminal tab
  • pm2 start dist/server.js --name mcp to survive terminal closes and auto-restart on crash

stdio mode does not need this — the agent spawns the process itself.

Adding a new tool

  1. Create src/tools/yourTool.ts exporting registerYourTool(server)
  2. Call it from src/tools/index.ts
  3. Keep the tool narrow - one job, a tight zod schema, a description that tells the model exactly when to use it and what format inputs should be in

Security notes

  • The HTTP server binds to 127.0.0.1 only - not reachable from outside this machine
  • DNS rebinding protection is applied automatically by @modelcontextprotocol/fastify
  • Set MCP_TOKEN to require a bearer token even for local HTTP calls
  • stdio transport is inherently authenticated by process spawn — no token needed
  • Keep total tools per server in the 6-10 range - Cursor caps out around 40 active tools across all connected servers combined
  • 127.0.0.1:8787/mcp
  • npx @modelcontextprotocol/inspector node dist/stdio-server.js

node dist/cli.js image.png

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选