pdf-extraction skill for Claude: pull text and tables out of PDFs with pdfplumber

Makes Claude write and run pdfplumber code to pull text, tables and metadata out of a PDF, then turn tables into CSV or Excel.

Installs: 9.4K; on skills.sh
Stars: 556; on GitHub
Updated: Jan 2026; last repo commit

You give Claude a PDF and say what you need: text, tables, images or metadata. Claude then writes pdfplumber code and runs it. The skill also covers cropping to a region, filtering by font, multi-column pages and a visual debug image of detected tables.

It works on PDFs with real text. SKILL.md says scanned or image PDFs need OCR first, complex layouts may need tuning, some encryption types are not supported, and it cannot edit PDFs. It needs Python and pdfplumber installed.

Good for

Analysts who need tables from financial reports or invoices
Developers who want a reusable pdfplumber workflow
Anyone with text-based PDFs, not scans

Install pdf-extraction

Claude Code

Copy this prompt.

Install the pdf-extraction skill in Claude Code. Run this command:

npx skills add https://github.com/claude-office-skills/skills --skill pdf-extraction

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

Pick one route. The skill needs Python and runs pdfplumber code, so install pdfplumber first: pip install pdfplumber (add Pillow for debug images, and pandas for the DataFrame examples). There is no plugin route in the repo and no Claude app upload route, because the skill runs local Python.

Built from the package data we hold. Not yet tested by us.

What's in the skill

When it loads

Extract text, tables, and metadata from PDFs using pdfplumber

Sections in SKILL.md

Overview
How to Use
Domain Knowledge
pdfplumber Fundamentals
PDF Structure
Text Extraction
Table Extraction
Visual Debugging
Cropping and Filtering
Metadata and Structure
Best Practices
Common Patterns
Extract All Tables to DataFrames
Extract Specific Region
Multi-column Layout
Examples

Key facts

Author: claude-office-skills (named in the SKILL.md front matter)
License: MIT (SKILL.md front matter; the README says individual skills may differ)
Installs on skills.sh: 9,391 (Nov 2, 2026)
GitHub stars: 556 for the whole claude-office-skills/skills repo (Nov 2, 2026)
Last repo commit: Jan 31, 2026
Audits on skills.sh: Gen Agent Trust Hub: pass, Socket: pass, Snyk: pass
What it includes: 1 file: SKILL.md (bundles no scripts; code is inside the instructions)
Needs: Python and pip install pdfplumber (optional Pillow for debug images)

What it includes

Pulls text, words and characters with positions
Examples for plain text, layout-preserving text, words with positions and character-level font and size.
Try: “Get text from pages 5 to 10 of this document.”
Runs code · Text extraction
Extracts tables with default or custom settings
Uses extract_tables, with table settings for line or text strategies and tolerances, and find_tables to locate tables without extracting.
Try: “Extract all tables from this financial report.”
Runs code · Table extraction
Saves images showing detected objects and tables
Uses page.to_image to draw rectangles and lines and to_image().debug_tablefinder() for tables, and saves PNG files.
Write · Visual debugging
Limits extraction to a region, font or size
Crops a page to a bounding box, filters characters by position, and extracts bold or large text.
Runs code · Crop and filter
Reads PDF metadata and page info
Prints title, author, creation date, page sizes and rotation.
Runs code · Metadata and structure
DataFrames, invoice totals, multi-column, resume parser
Examples turn tables into pandas DataFrames, find an invoice total near the word Total, split columns, parse a resume, and write an Excel workbook with one sheet per table.
Try: “Find and extract the invoice total from this PDF.”
Runs code · Common patterns and examples
Scanned PDFs, complex layouts, encryption, editing
SKILL.md lists: no scanned or image PDFs without OCR, complex layouts may need tuning, some encryption not supported, embedded fonts can affect text, no PDF editing.
Read · Limitations

From the official docs

Permissions

Medium risk

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

It runs generated Python code
SKILL.md says Claude will generate pdfplumber code and execute it. Run it on PDFs you trust and review the code if the files are sensitive. github.com
It installs packages
The Installation section lists pip install pdfplumber and optionally Pillow. The examples also import pandas, which that section does not list. Use a virtual environment. github.com
Examples write files to your folder
The financial-report example saves a debug PNG per page and extracted_tables.xlsx in the working directory. Run it in a scratch folder. github.com
The front matter names an MCP server
The metadata lists an office-mcp server with extract_text_from_pdf and extract_tables_from_pdf. The instructions themselves write pdfplumber code and do not call those tools. github.com

Security advisories

No published security advisories as of Nov 2, 2026.

Technical details

Turning on read-only	No read-only mode. Use version control.
Sign-in and keys	No key or account.
Scopes	Claude writes Python and runs it on your machine against PDFs you give it Installs Python packages with pip Example code writes PNG debug images and Excel files

Questions

What does pdf-extraction do?

It makes Claude write and run pdfplumber code to extract text, tables and metadata from PDFs. It can also crop to a region, filter by font and save tables as CSV or Excel.

How do I install it?

Use the install box at the top of this page, then install pdfplumber with pip. The skill is one file and does not install Python packages itself.

Does it read scanned PDFs?

No. SKILL.md says it is for native text and that scanned or image PDFs need OCR first.

Does it need an API key?

SKILL.md names no key. It uses local Python libraries. The front matter mentions an office-mcp server, but the instructions write pdfplumber code instead.

Is it safe?

Claude writes Python and runs it on your PDFs, installs packages with pip and may save images and Excel files in your folder. Use a scratch folder and a virtual environment, and avoid running it on files you do not trust.

Who made it?

The SKILL.md front matter names claude-office-skills as author and MIT as license. The repo is a collection of office skills; this one is its pdf-extraction folder.

Information

updated Nov 2, 2026

Repository: claude-office-skills/skills
Folder: pdf-extraction

claude-office-skills/skills skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 2

pdf-extraction SKILL.md · github.com
claude-office-skills README · github.com