document-ingestion

Convert untrusted PDFs and Office documents into verified Markdown evidence packs.

Updated Jul 19, 2026

Search Tags:#markdown #research #docx #document ingestion #pdf conversion #markitdown #evidence extraction

Ready to run

npx skills add https://github.com/DagimAlemayehuu/AgenticEngineering --skill document-ingestion-dagimalemayehuu

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: document-ingestion
Source: https://github.com/DagimAlemayehuu/AgenticEngineering/tree/main/agentic-engineering/core/dependencies/research/document-ingestion
Command: npx skills add https://github.com/DagimAlemayehuu/AgenticEngineering --skill document-ingestion-dagimalemayehuu

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires markitdown, and includes references (resource) components.

What problem does it solve? External documents like PDFs, slides, and spreadsheets cannot be safely or reliably fed into research workflows without conversion, fidelity checks, and controls against treating document content as instructions. ## Core Features & Use Cases - Deterministic Conversion: Uses MarkItDown in an isolated environment to convert PDF, DOCX, PPTX, XLSX, HTML, CSV, JSON, and XML files into Markdown. - Fidelity Verification: Compares page counts, headings, tables, and critical passages against the original before the output is trusted, flagging scanned or table-heavy documents as partial. - Safe Evidence Packs: Extracts only relevant excerpts with source locators and citations for NotebookLM, Jules, or local agents, never forwarding entire documents by default. - Use Case: Before attaching a vendor whitepaper PDF to an engineering research task, convert it locally, verify the extraction, and supply only a cited evidence pack to the downstream agent. ## Quick Start Use $document-ingestion to convert this document into verified Markdown evidence.

Frequently Asked Questions about document-ingestion

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

FAQPage Schema

How do I convert a PDF to Markdown with MarkItDown?▼

Install MarkItDown in a project-local virtual environment with the needed format extras, then run markitdown input.pdf -o evidence.md. Verify the output against the original before relying on it for engineering decisions.

What file formats does MarkItDown support for conversion?▼

MarkItDown supports PDF, DOCX, PPTX, XLSX, HTML, CSV, JSON, XML, images, audio, EPUB, and ZIP archives. Install only the format extras you need, such as markitdown[pdf,docx,pptx].

Does MarkItDown work on scanned PDFs or complex tables?▼

MarkItDown handles text-native documents well, but scanned PDFs, diagrams, complex columns, and legal text need a visual or manual check. Mark such extractions as partial and use an approved OCR fallback if conversion is incomplete.

Should I enable MarkItDown OCR or LLM plugins?▼

Only with explicit approval. OCR and LLM plugins can send document data to external models and incur provider charges, so they require sign-off on the provider, cost, and data handling before use.

How do I safely use external documents with AI agents?▼

Treat every document as untrusted data, never as instructions. Preserve the original, create a verified Markdown derivative, and supply only a compact evidence pack with citations rather than the full document.