historical-document-ocr
An accurate OCR tool for transcribing faded, handwritten, or legacy document scans into digital text.
Install
mkdir -p .claude/skills/historical-document-ocr && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/12312" && unzip -o skill.zip -d .claude/skills/historical-document-ocr && rm skill.zipInstalls to .claude/skills/historical-document-ocr
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Transcribe scanned/photographed historical documents (PDFs or images) to text with Gemini vision, using accuracy-first practices: per-page high-res rendering, faded-scan image enhancement, strict verbatim prompting, and an optional multi-model consensus pass that reconciles disagreements by re-reading the page. Use this whenever the user wants to OCR, transcribe, or extract the text of scanned letters, manuscripts, typescripts, carbon copies, ledgers, archival records, genealogy documents, old correspondence, or any image-only PDF that has no real text layer — especially when the material is handwritten, typewritten, faded, rotated, or hard to read and accuracy matters. Trigger even if the user just says "read this old scan", "what does this letter say", "digitize these archive pages", or "transcribe this PDF" and the PDF turns out to be scanned images. Do NOT use this for audio/podcast transcription (that is gemini-podcast-transcribe) or for born-digital PDFs that already contain selectable text (a plain pdftotext is enough there).Key capabilities
- →Transcribe scanned historical documents to text
- →Render high-resolution pages for review
- →Enhance faded-scan images
- →Use context blocks to disambiguate ambiguous characters
- →Perform multi-model consensus pass for accuracy
- →Report results and review list for human verification
How it works
The skill transcribes scanned historical documents using Gemini vision, optimizing for accuracy on degraded material. It renders pages, enhances images, uses context for disambiguation, and can perform a multi-model consensus pass.
Inputs & outputs
When to use historical-document-ocr
- →Transcribing historical archives
- →Converting scanned PDFs to text
- →Digitizing ledger pages
- →Reading handwritten manuscripts
About historical-document-ocr
Transcribes scanned images or PDFs using AI vision models. It is specifically optimized for high accuracy on degraded material like handwritten notes or faded manuscripts.
>-
When not to use it
- →For audio/podcast transcription
- →For born-digital PDFs that already contain selectable text
- →When `pdftotext` already returns the real text
Prerequisites
Limitations
- →Expects ~1–4% word error on faded/handwritten material
- →Does not invent text; failure modes are skipped lines and word substitutions
- →Requires human pass on flagged disagreement spans for archival-grade output
How it compares
This skill prioritizes accuracy on degraded historical documents through image enhancement, context-aware disambiguation, and multi-model consensus, unlike standard OCR tools that may struggle with such material.
Compared to similar skills
historical-document-ocr side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| historical-document-ocr (this skill) | 0 | 4mo | Review | Intermediate |
| scientific-paper | 0 | 5mo | No flags | Intermediate |
| content-research-writer | 15 | 11mo | No flags | Beginner |
| research-grants | 6 | 9mo | Review | Advanced |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by tdhopper
View all by tdhopper →You might also like
scientific-paper
cepdnaclk
Use this skill when writing, revising, or checking the FYP research paper in this repository.
content-research-writer
ComposioHQ
Assists in writing high-quality content by conducting research, adding citations, improving hooks, iterating on outlines, and providing real-time feedback on each section. Transforms your writing process from solo effort to collaborative partnership.
research-grants
davila7
Write competitive research proposals for NSF, NIH, DOE, and DARPA. Agency-specific formatting, review criteria, budget preparation, broader impacts, significance statements, innovation narratives, and compliance with submission requirements.
evidence-draft
WILLOSCAR
|
figure-caption
Taichuan3
用于生物信息学论文图表规划、figure title、panel title、legend、caption、figure-to-claim 审查和图注中的 source data/caveat 表达。
linkedin-sales-navigator-alt
OneWave-AI
Build targeted prospect lists by analyzing LinkedIn profiles, extracting job titles, companies, locations, and recent activity. Identifies decision-makers, tracks job changes for warm outreach, and enriches contact data. Use when users need to find prospects, build lead lists, or track decision-maker movements.