# PDF to Markdown review checklist

Reviewed: 2026-09-02  
Source: https://pdfmd.org/blog/pdf-to-markdown-guide

Use this after conversion and before publishing, calculating, citing, or feeding the result into another system.

## Document route

- [ ] I confirmed whether the PDF has a usable text layer or needs OCR.
- [ ] I checked beyond page one for scanned exhibits or mixed page types.
- [ ] I know which pages used text conversion and which used OCR.

## Reading order and headings

- [ ] I compared the first and last block on every complex page with the PDF.
- [ ] I checked columns, sidebars, captions, and footnotes for interleaving.
- [ ] Heading levels form one coherent outline without skipped context.

## Tables and formulas

- [ ] Header and delimiter rows have the expected number of columns.
- [ ] I checked the first, middle, and last row of every important table.
- [ ] Merged cells, repeated headers, units, footnotes, and cross-page rows were reviewed manually.
- [ ] Every formula and symbol used downstream was checked against the source.

## Provenance and export

- [ ] Important claims retain a page reference or source link.
- [ ] I did not treat missing coordinates or confidence as evidence.
- [ ] I selected the export profile for the destination: clean Markdown, Obsidian, GitHub/MkDocs, RAG JSONL, or source-map JSON.
- [ ] I downloaded or copied the result before leaving; pdfmd does not save it automatically.

## Before reuse

- [ ] I removed OCR artifacts and repaired words split by line endings.
- [ ] I checked names, dates, totals, negative signs, decimal separators, and units.
- [ ] A human reviewed any value that will drive a decision, calculation, or public claim.

Product and evidence changes: https://pdfmd.org/changelog