PDF Reader

Reads PDFs, extracts text/images/tables, and provides resources to AI agents for processing

python

0

GitHub Stars

python

Language

6 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "rexfelix-readpdf_mcp_server": {
      "command": "uv",
      "args": [
        "--directory",
        "/YOUR/ABSOLUTE/PATH/TO/pdf_reader_mcp_server",
        "run",
        "src/server/main.py"
      ],
      "env": {
        "ENV": "YOUR_VALUE"
      }
    }
  }
}

You can install and run the PDF Reader MCP Server to extract text, images, and tables from PDFs and feed them to an AI agent. This server makes PDF data accessible through a simple, script-based interface so you can automate reading and summarizing documents.

How to use

Use an MCP client to connect to the PDF Reader MCP Server and call its tools to process PDFs. The server exposes tools such as read_pdf to extract content as Markdown, and get_pdf_metadata to inspect metadata like title and page count. You can also request a direct resource read via the pdf:// URI pattern. When you issue a request, provide the source path or URL to the PDF and, if needed, a page range. For example, you can ask the agent to summarize a PDF after reading it to get a concise overview.

Available tools and usage patterns

read_pdf reads the PDF contents and returns Markdown text. Typical inputs include source pointing to the PDF and an optional page_range like "1-3" to limit the extraction to specific pages.
get_pdf_metadata retrieves metadata such as title and page count. Use inputs like source set to the PDF URL or file path.
Access PDFs directly via URI patterns, for example: pdf://{absolute_path_to_file}. You can prompt the AI to read a file with a request like 이 파일 읽어줘: pdf:///Users/me/docs/manual.pdf, which triggers internal resource reading.

Prompts and features

You can use predefined prompts such as summarize_pdf to have the server read a specified PDF and produce a concise summary.

Available tools

read_pdf

Reads the PDF contents and returns Markdown text. Accepts inputs like source for the file path and optional page_range to limit extraction.

get_pdf_metadata

Fetches metadata such as title and page count from the PDF source.

read_resource

Reads a resource directly from a given URI pattern like pdf://{absolute_path} and returns its content to the AI agent.

summarize_pdf

Reads the specified PDF and produces a concise summary based on its content.