Can I customize how the text is extracted?

Yes, you can provide custom instructions with the command, such as 'extract only tables' or 'convert this to English,' and the vision model will follow those instructions during conversion.

Which file formats does the OCR skill support?

The skill supports PDF files and various image formats including .png, .jpg, .jpeg, .webp, .gif, .bmp, and .tiff.

How does this skill handle large documents without hitting context limits?

It uses a Task agent to perform the OCR and file writing operations. This ensures that the heavy file content stays within the subagent's scope, returning only the status to your main conversation.

Can I process multiple files at once?

Yes. If you provide a folder path, the skill enters Batch Mode and processes all supported files in parallel groups of three for speed and efficiency.

OCR to Markdown

Name: OCR to Markdown
Author: Bae-ChangHyun

byBae-ChangHyun

•

Content Management

Converts PDFs and images into structured Markdown documents using Claude's vision capabilities while preserving main context efficiency.

This skill enables Claude to perform high-fidelity Optical Character Recognition (OCR) on PDFs and a wide range of image formats, transforming them into cleanly formatted Markdown. By utilizing dedicated subagents for processing, the skill protects the main conversation from context bloat and token overflow, allowing for the extraction of tables, code blocks, and complex document structures without performance degradation. Whether processing a single screenshot or a folder full of multi-page reports, it provides flexible output options including unified file generation or page-by-page exports.

Key Features

01Parallel batch processing for entire folders in groups of three

021 GitHub stars

03Subagent-based execution to prevent main context memory exhaustion

04Structure-aware conversion that preserves tables, headers, and code blocks

05Multi-format support for PDFs and images including PNG, JPG, WEBP, and TIFF

06Flexible PDF output modes for unified or page-by-page Markdown generation

Use Cases

01Extracting data from screenshots of UI or code snippets directly into a project

02Converting scanned technical documentation or whitepapers into editable Markdown files

03Automating the digitizing of large batches of PDF reports with custom extraction rules

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bae-changhyun/cc-plugins-bch ocr

For use in Claude.ai and ChatGPT

Download Skill