How do I install the Ocr And Documents skill?

Run npx skillmds@latest add ekkolearnai/ocr-and-documents in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Ocr And Documents safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Ocr And Documents?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Ocr And Documents free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Ocr And Documents?

ekkolearnai (@ekkolearnai) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Ocr And Documents by ekkolearnai — Agent Skill

Name: Ocr And Documents
Author: ekkolearnai

File contents

OCR and scanned documents

Use this Skill when ordinary document extraction fails because the source is scanned, image-based, or layout-heavy.

Start here

Call skill_view for this Skill and use its returned baseDirectory. Helpers live in /scripts.

Prefer local processing. Do not upload a private document to an external OCR service unless the user explicitly authorizes that transfer.

Choose the lightest path

Normal extraction first. For PDFs, try the bundled pdf Skill. For .docx, use docx. Do not OCR born-digital text.
A few scanned pages. Render pages with the pdf Skill and inspect images using view_image; transcribe only the requested pages.
Text-oriented scanned PDF. Try extract_pymupdf.py; this is the lighter local path.
Complex layout, tables, math, or many scanned pages. Use extract_marker.py only after confirming the need and checking resources.

Helpers

python3 <baseDirectory>/scripts/extract_pymupdf.py input.pdf -o extracted.md
python3 <baseDirectory>/scripts/extract_marker.py --check
python3 <baseDirectory>/scripts/extract_marker.py input.pdf -o extracted.md

Run helpers with terminal_exec, using python3 as the command and arguments as an array. Use --help for exact options.

extract_pymupdf.py requires PyMuPDF:

python3 -m pip install pymupdf

Its --markdown mode additionally requires pymupdf4llm.

extract_marker.py uses marker-pdf, which can require several gigabytes of packages and model downloads. Never install it automatically. Check available disk space and ask the user before installing or downloading large models.

URLs

If the source is a URL, first use an available browser or HTTP tool to identify whether it points to HTML or a document. Download only the requested document, respect authentication and access boundaries, and then process the local copy. Do not assume a specific web tool exists.

Verification

Preserve reading order, page references, headings, tables, footnotes, and confidence caveats.
Compare extracted text against representative rendered pages with view_image.
Mark uncertain characters and fields instead of guessing.
For tables, verify column alignment and numeric values separately.
Return both the extracted artifact path and the original source/page mapping when possible.

---name: ocr-and-documentsdescription: Extract and structure text from scanned PDFs and document images with a local-first OCR workflow, layout-aware fallbacks, and explicit heavyweight dependency controls.---# OCR and scanned documentsUse this Skill when ordinary document extraction fails because the source is scanned, image-based, or layout-heavy.## Start hereCall skill_view for this Skill and use its returned baseDirectory. Helpers live in <baseDirectory>/scripts.Prefer local processing. Do not upload a private document to an external OCR service unless the user explicitly authorizes that transfer.## Choose the lightest path1. Normal extraction first. For PDFs, try the bundled pdf Skill. For .docx, use docx. Do not OCR born-digital text.2. A few scanned pages. Render pages with the pdf Skill and inspect images using view_image; transcribe only the requested pages.3. Text-oriented scanned PDF. Try extract_pymupdf.py; this is the lighter local path.4. Complex layout, tables, math, or many scanned pages. Use extract_marker.py only after confirming the need and checking resources.## Helpers python3 <baseDirectory>/scripts/extract_pymupdf.py input.pdf -o extracted.md python3 <baseDirectory>/scripts/extract_marker.py --check python3 <baseDirectory>/scripts/extract_marker.py input.pdf -o extracted.mdRun helpers with terminal_exec, using python3 as the command and arguments as an array. Use --help for exact options.extract_pymupdf.py requires PyMuPDF: python3 -m pip install pymupdfIts --markdown mode additionally requires pymupdf4llm.extract_marker.py uses marker-pdf, which can require several gigabytes of packages and model downloads. Never install it automatically. Check available disk space and ask the user before installing or downloading large models.## URLsIf the source is a URL, first use an available browser or HTTP tool to identify whether it points to HTML or a document. Download only the requested document, respect authentication and access boundaries, and then process the local copy. Do not assume a specific web tool exists.## Verification- Preserve reading order, page references, headings, tables, footnotes, and confidence caveats.- Compare extracted text against representative rendered pages with view_image.- Mark uncertain characters and fields instead of guessing.- For tables, verify column alignment and numeric values separately.- Return both the extracted artifact path and the original source/page mapping when possible.

Frequently asked questions

Ocr And Documents