Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@PDF Extraction MCP Server extract text from the PDF at /documents/report.pdf"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP server to extract contents from PDF files, with fixes for Claude Code CLI installation.
This fork includes critical fixes for installing and running the server with Claude Code (the CLI version).
Added __main__.py - Enables the package to be run as a module with python -m pdf_extraction
Claude Code specific instructions - Clear installation steps that work with Claude Code CLI
Tested installation process - Verified working with claude mcp add command
Related MCP server: mcp-pdf-tools
The server implements one tool:
extract-pdf-contents: Extract contents from a local PDF file
Takes pdf_path as a required string argument (local file path)
Takes pages as an optional string argument (comma-separated page numbers, supports negative indexing like -1 for last page)
Supports both PDF text extraction and OCR for scanned documents
Python 3.11 or higher
pip or conda
Claude Code CLI installed (claude command)
# Clone this fork
git clone https://github.com/lh/mcp-pdf-extraction-server.git
cd mcp-pdf-extraction-server
# Install in development mode
pip install -e .# Check where pdf-extraction was installed
which pdf-extraction
# Example output: /opt/homebrew/Caskroom/miniconda/base/bin/pdf-extraction# Add the server using the full path from above
claude mcp add pdf-extraction /opt/homebrew/Caskroom/miniconda/base/bin/pdf-extraction
# Verify it was added
claude mcp list# Start a new Claude session
claude
# In Claude, type:
/mcp
# You should see:
# MCP Server Status
# • pdf-extraction: connectedOnce connected, you can ask Claude to extract PDF contents:
"Can you extract the content from the PDF at /path/to/document.pdf?"
"Extract pages 1-3 and the last page from /path/to/document.pdf"Make sure you started a NEW Claude session after adding the server
Verify the command path is correct: ls -la $(which pdf-extraction)
Test the command directly (it should hang waiting for input): pdf-extraction
If you get Python import errors:
Make sure you're using the same Python environment where you installed the package
Try using the full Python path: claude mcp add pdf-extraction /path/to/python -m pdf_extraction
If pip install -e . fails:
Make sure you have Python 3.11+: python --version
Try creating a fresh virtual environment:
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -e .This fork is specifically for Claude Code CLI. If you're using Claude Desktop (the GUI app), please refer to the original repository for installation instructions.
mcp>=1.2.0
pypdf2>=3.0.1
pytesseract>=0.3.10 (for OCR support)
Pillow>=10.0.0
pydantic>=2.10.1,<3.0.0
pymupdf>=1.24.0
Contributions are welcome! The main change in this fork is the addition of __main__.py to make the package runnable as a module.
Same as the original repository.
Extract contents from a local PDF file, given page numbers separated in comma. Negative page index number supported.
| Name | Required | Description | Default |
|---|---|---|---|
| pdf_path | Yes | ||
| pages | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It states the tool extracts contents but doesn't disclose behavioral traits such as what format the extracted contents are in (e.g., text, images), error handling, permissions needed for local file access, or performance considerations. The mention of negative page indices adds some context but is insufficient for a mutation-like operation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is brief and to the point with two sentences, front-loading the main action. It avoids unnecessary words, though it could be slightly more structured (e.g., separating parameter explanations).
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given no annotations, 0% schema coverage, and no output schema, the description is incomplete. It lacks details on return values, error cases, file format support, and operational constraints, making it inadequate for a tool that interacts with local files and performs extraction.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate. It explains that 'pages' accepts 'page numbers separated in comma' and supports 'negative page index number', which adds meaning for one parameter. However, it doesn't clarify 'pdf_path' (e.g., file path format, supported locations) or other details, leaving gaps for the 2 parameters overall.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the verb 'extract' and the resource 'contents from a local PDF file', making the purpose understandable. It specifies the scope of extraction (by page numbers) but doesn't distinguish from siblings since none exist, so it cannot achieve a perfect score of 5.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives, prerequisites, or exclusions. It mentions 'negative page index number supported' which hints at a feature but doesn't clarify usage context or limitations.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
With only one tool, there is no possibility of ambiguity or overlap with other tools. The tool's purpose is clearly defined as extracting contents from PDF files, making it impossible for an agent to confuse it with another tool.
Since there is only one tool, naming consistency is inherently perfect. The tool name 'extract-pdf-contents' follows a clear verb-noun pattern, and there are no other tools to deviate from or conflict with this pattern.
A single tool for a PDF extraction server is too minimal for the apparent scope. While the tool handles content extraction, typical PDF operations like metadata retrieval, text analysis, or format conversion are missing, making the server feel incomplete and underpowered for its domain.
The server is severely incomplete for PDF extraction. It only provides content extraction with page selection, lacking essential operations such as getting metadata, extracting images, converting to other formats, or searching within PDFs. This will likely cause agent failures when broader PDF tasks are required.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
MCP server for detecting and redacting PII (Personally Identifiable Information) in PDF documents.
Hosted MCP server: convert PDFs to clean, LLM-ready Markdown with tables, formulas and OCR.
MCP server for the PDFGate API. Generate PDFs, manage documents and handle e-signatures.
MCP server for Api2Pdf — generate PDFs & images from HTML, URLs or office files; merge, barcodes.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/xraywu/mcp-pdf-extraction-server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
