# PdfPecker MCP server

> Converts HTML and Markdown to PDF and extracts text, images, or Markdown from existing PDF documents.

- Repository: https://github.com/Moonwalker1307/PdfPecker.Python
- Page: https://hostedmcp.org/servers/moonwalker1307/pdfpecker-python
- Transport: stdio
- Stars: 1
- License: MIT
- Latest version: 1.25.0

## Install

### Claude Code

```sh
claude mcp add pdfpecker-python -- uvx pdfpecker
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "pdfpecker-python": {
      "command": "uvx",
      "args": [
        "pdfpecker"
      ]
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "pdfpecker-python": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "pdfpecker"
      ]
    }
  }
}
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "pdfpecker-python": {
      "command": "uvx",
      "args": [
        "pdfpecker"
      ]
    }
  }
}
```

### Codex (`~/.codex/config.toml`)

```toml
[mcp_servers.pdfpecker-python]
command = "uvx"
args = ["pdfpecker"]
```

### Gemini CLI (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "pdfpecker-python": {
      "command": "uvx",
      "args": [
        "pdfpecker"
      ]
    }
  }
}
```

## FAQ

### How does PdfPecker handle images and CSS in HTML files?

Images, stylesheets, and CSS resolve relative to the file's folder using the base_dir option.

### Is the server safe for AI agents to use in a sandbox?

Yes, it is designed for sandboxed environments as it requires only one pip install, has no runtime downloads, and restricts file access to specified root folders.