verify-translation

Compare translated text against PDF page images to detect extraction gaps.

Updated Jan 2, 2026
One-click install
npx skills add https://github.com/Takazudo/takazudomodular-manuals --skill verify-translation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verify-translation
Source: https://github.com/Takazudo/takazudomodular-manuals/tree/main/.claude/skills/verify-translation
Command: npx skills add https://github.com/Takazudo/takazudomodular-manuals --skill verify-translation

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires playwright, and includes scripts (resource) components.

What problem does it solve?

This Skill provides automated verification that translations align with the original PDF page images, detecting missing content and ensuring fidelity across bilingual manuals.

Core Features & Use Cases

  • Image-to-text comparison: captures high-resolution page screenshots and compares the translated text against the corresponding PDF image.
  • Extraction-failure detection: identifies pages where text extraction fell short or is incomplete.
  • Regeneration workflow: can regenerate extracted text from images to support re-translation and re-verification.
  • Use Case: After completing translation, run this skill to generate a correctness report and flag pages needing manual fixes before deployment.

Quick Start

Run the verification script against a slug to capture all pages and produce a verification summary: node .claude/skills/verify-translation/scripts/capture-pages.js --slug oxi-e16-manual --pages 74

Frequently Asked Questions about verify-translation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify translated text against original PDF page images?

You can verify translated text against original PDF page images by running a screenshot-based comparison workflow that captures high-resolution page screenshots and compares translated content against corresponding PDF images to detect missing content.

What is the best way to detect extraction gaps in bilingual manual translations?

The best way to detect extraction gaps in bilingual manual translations is using an image-to-text comparison that captures page screenshots and identifies pages where text extraction fell short or is incomplete.

Do I need Playwright to run the translation verification script?

Yes, you need Playwright to run the translation verification script because it is a required dependency for capturing the page screenshots used in the image comparison workflow against the production build of the manual viewer.

How do I regenerate missing text after translation fails verification?

To regenerate missing text after translation fails verification, use the regeneration workflow to extract text from the original page images, which supports re-translation and subsequent re-verification of the bilingual manual.

What limitations exist when verifying translations against PDF images?

A key limitation when verifying translations against PDF images is that the process requires a production build of the manual viewer, access to page manifests, and existing translations to operate the screenshot-based comparison workflow.