pdf-extraction

Convert PDF documents into enriched Markdown with tables, image descriptions, and pseudocode.

Updated Apr 20, 2026

Search Tags:#pdf #markdown #mcp #table extraction #pseudocode #vlm #document enrichment

Ready to run

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

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pdf-extraction
Source: https://github.com/thistleknot/skills/tree/main/pdf-extraction
Command: npx skills add https://github.com/thistleknot/skills --skill pdf-extraction-thistleknot

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It eliminates the manual effort of converting PDFs into clean, structured documents by extracting text, tables, image descriptions, and method-level pseudocode into a single enriched Markdown artifact.

Core Features & Use Cases

Full PDF → enriched Markdown pipeline: orchestrates a multi-phase conversion that produces both document text and value-added method pseudocode.
Table enhancement and fusion: improves tabular accuracy by combining multiple extractors and using a VLM-based arbitration step.
VLM image descriptions + reinsertion: generates descriptions for extracted images and embeds them back into the Markdown.
Method extraction for RAG: produces a {stem}_methods.md output focused on algorithmic contribution pseudocode.

Quick Start

Run run_pipeline.bat with your PDF path to generate the enriched Markdown and the companion _methods.md file beside it.

Frequently Asked Questions about pdf-extraction

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I convert PDF documents with complex tables into Markdown for RAG?▼

To convert PDFs into Markdown for RAG, this pipeline fuses extracted tables using multiple extractors and VLM-based arbitration to ensure tabular accuracy. It outputs a structured Markdown artifact with enhanced tables ready for downstream chunking.

Can I extract text and image descriptions from PDFs automatically?▼

Extracting text and image descriptions from PDFs is handled by applying VLM-generated descriptions to extracted images and embedding them back into the Markdown. This process eliminates manual effort by producing a single enriched document artifact.

What is the best way to extract method-level pseudocode from a PDF?▼

The best way to extract method-level pseudocode from a PDF is using the GPT-based method extraction step in the pipeline. It processes the document to produce a separate Markdown file focused on algorithmic contributions for RAG workflows.

Does PDF to Markdown conversion work with annotated figures and complex layouts?▼

PDF to Markdown conversion works with annotated figures and complex layouts through a multi-phase pipeline using docling-based base extraction. It applies post-processing to image tokens and reinserts VLM descriptions to handle arbitrary document structures.

Do I need a VLM to improve table extraction accuracy from PDF documents?▼

A VLM is required to improve table extraction accuracy from PDF documents because it acts as an arbitration step to fuse results from multiple extractors like tabula and camelot. This resolves conflicts and ensures clean structured output.