| name | research-reference-pdf-txt-conversion |
| description | Converts research PDFs into readable TXT mirrors using text extraction first and OCR fallback second. Use when handling scanned literature, checking text-layer quality, improving reference coverage, or generating reviewable OCR manifests. |
Research Reference PDF-TXT Conversion
Purpose
Use this skill when building or improving a literature library for research or writing.
Typical tasks:
- convert PDFs to TXT mirrors
- detect whether PDFs already contain text
- route scanned files to OCR
- generate manifests and quality reports
- improve coverage in small reviewable batches
Core rules
- Audit first, then apply.
- Prefer direct text extraction before OCR.
- Do not overwrite good TXT files unless explicitly requested.
- Mark poor results honestly as needs_ocr or failed.
- Always leave a report or candidate board.
Suggested flow
Route A — Text-layer audit
Check whether the PDF is already machine-readable.
Route B — Direct extraction
Extract usable text and generate a TXT mirror.
Route C — OCR fallback
Only use OCR for scanned or very low-text files.
Route D — Batch review
Review quality scores before scaling the next batch.
Expected outputs
- TXT mirrors
- manifest files
- OCR candidate board
- conversion summary report
Boundary
This skill stops at conversion and audit. Evidence extraction, literature synthesis, and paragraph writing should be handled by separate skills.
