AGENT SKILL

pdf-ocr-local

Extract text from scanned or image-heavy PDFs using available local OCR/PDF tooling, then preserve page references for downstream analysis.

License 许可
MIT
最近更新
2026-07-28
Stars
★ 40
Forks
1

01 前言 Front Matter

name
pdf-ocr-local
description
Extract text from scanned or image-heavy PDFs using available local OCR/PDF tooling, then preserve page references for downstream analysis.

02 资源文件 3

03 SKILL.md 正文

Local PDF OCR Skill

Use the source documents already available in the workspace. Do not request external OCR credentials. Use local tools that are already available, such as pdftotext, Python PDF libraries, image conversion tools, or OCR binaries if present.

Workflow

  1. Inventory the PDF pages and identify text-native versus scanned/image-heavy pages.
  2. Extract text with available local tools.
  3. For weak OCR pages, record the page number and the confidence/uncertainty in your notes.
  4. Preserve page references in the final packet so the user can verify important facts.
  5. When the user requests Word/PDF output, use available document tooling and validate that required sections are present.

Never expose credentials in outputs.

04 同仓库更多 Skills