A specialized skill for translating PDFs locally by extracting text and leveraging agent-native language capabilities for high-fidelity output.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install pdf-translation-reserving-layout
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install pdf-translation-reserving-layout using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is PDF Translate BabelOCR?
PDF Translate BabelOCR is a developer-centric skill designed to handle the complexities of translating born-digital or lightly scanned PDFs within the Openclaw Skills ecosystem. Unlike traditional cloud-based translators, this skill prioritizes local text extraction and utilizes the agent's internal language processing power to generate translated Markdown, JSON, or plain text. It is specifically engineered to handle research papers and technical documents where maintaining reading order and structure is critical.
By focusing on a local-first workflow, PDF Translate BabelOCR ensures data privacy and allows for high-granularity control over the translation process. It provides a bridge between raw PDF data and structured translated content, making it an essential addition for users of Openclaw Skills who require repeatable, high-quality document processing without relying on external LLM services as a primary driver.
PDF Translate BabelOCR Use Cases
- Translating academic research papers while preserving page-specific context.
- Processing large technical manuals by parallelizing translation across multiple subagents.
- Converting non-English PDF documentation into structured Markdown for documentation sites.
- Extracting text from digitally scrambled PDFs that traditional OCR might fail to parse correctly.
How PDF Translate BabelOCR Works
- The skill utilizes a bundled Python extractor to pull text page-by-page from the source PDF using layout-aware tools.
- Extracted content is stored in a JSONL format, mapping every block of text to its original page number.
- A batching helper analyzes the JSONL file and splits the content into manageable chunks based on character and page limits.
- The agent or assigned subagents translate these batches, maintaining formatting constraints such as headings and table labels.
- The skill merges the translated batches back into a cohesive document, performing consistency checks across batch boundaries.
PDF Translate BabelOCR Setup
To get started with this skill in your environment, ensure you have the necessary scripts available and run the following extraction command:
python3 scripts/extract_pdf_pages.py \
--input /absolute/path/document.pdf \
--output /absolute/path/work/pages.jsonl
For larger documents, use the batching helper to prepare the files for the translation phase within Openclaw Skills:
python3 scripts/build_translation_batches.py \
--input /absolute/path/work/pages.jsonl \
--output-dir /absolute/path/work/batches \
--max-pages 8 \
--max-chars 18000
PDF Translate BabelOCR Data Schema & Taxonomy
The skill organizes its data pipeline using a clear metadata taxonomy to ensure reliability:
| File Type | Description | Key Fields |
|---|---|---|
| Input PDF | The source document for translation | N/A |
| pages.jsonl | Intermediate extraction file | page, text, layout_metadata |
| Batches | Split JSON files for parallel work | batch_id, page_range, content |
| Final Output | The translated result | page, source_text, translated_text |
PDF Translate BabelOCR Advanced Features
- Parallel translation support for distributing heavy workloads across disjoint subagents or workers.
- Intelligent batching logic specifically tuned for dense academic PDFs and complex layouts.
- Fallback integration for RTL (Right-to-Left) languages and heavily scanned documents when local extraction is insufficient.
- Automated consistency checks for terminology across page boundaries in long-form translations.
- Built-in support for preserving formula text and figure labels within the Openclaw Skills workflow.
SKILL.md
Loading
Related Openclaw Skills
Execute Openclaw Pipeline
An autonomous engine for programmatic SEO and AEO content generation that automates the full lifecycle from keyword research to WordPress publishing.
TTS Audio Mastering
A professional-grade workflow for refining synthetic text-to-speech audio into production-ready assets through cleanup, normalization, and timing alignment.
Word Document Manipulation with python-docx
A robust technical skill for automating Word document generation and template manipulation using python-docx while solving common XML run issues.
NotiLens
NotiLens is a real-time notification engine that delivers instant task lifecycle events, error alerts, and performance metrics from your AI agents directly to your phone.
von Neumann Mentor
An AI persona skill that adopts the identity of John von Neumann to teach computer science principles and software architecture through mathematical intuition and witty storytelling.
Git Assistant
An AI-powered git workflow tool that automates the generation of Conventional Commits, PR descriptions, and release documentation.
Featured*
