Extract and structure data from PDFs, Word docs, and emails using unified document processing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install data-extractor
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 data-extractor 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 Data Extractor?
The Data Extractor skill provides a unified interface for processing diverse document formats using the unstructured library. It is designed to transform raw files—ranging from scanned PDFs and Word documents to emails and HTML—into clean, structured data for downstream applications. This skill is a core asset within Openclaw Skills for developers who need to ingest messy data into vector databases or analytical pipelines.
By automatically detecting file types and applying intelligent partitioning strategies, it ensures consistent output regardless of the source. It handles the heavy lifting of OCR, table detection, and metadata extraction, allowing you to focus on building value rather than writing regex for every new document format.
Data Extractor Use Cases
- Converting research papers into structured JSON for automated literature reviews.
- Extracting line items and metadata from invoices and financial statements.
- Preparing unstructured legacy documentation for RAG-based search engines.
- Batch processing email archives to extract body content and attachments.
How Data Extractor Works
- Accept a document input or path within the Openclaw Skills workflow.
- Trigger the auto-partitioning engine to identify the file format and extract core elements.
- Apply OCR or high-resolution strategies for complex layouts or image-based files.
- Enrich extracted elements with granular metadata including page numbers and document coordinates.
- Output the structured elements into JSON, dictionaries, or dataframes for integration.
Data Extractor Setup
To install the base library for Openclaw Skills development, use:
pip install unstructured
For full document support including OCR and complex PDF parsing, use the all-docs bundle:
pip install "unstructured[all-docs]"
Data Extractor Data Schema & Taxonomy
The skill organizes data into a collection of Elements. Each element follows a predictable structure:
| Field | Type | Description |
|---|---|---|
| text | String | The raw text extracted from the document part. |
| category | String | The classification (e.g., Title, NarrativeText, Table, ListItem). |
| metadata | Object | Contextual data such as page_number, filename, and coordinates. |
| id | String | A unique hash identifying the specific element. |
Tables include an additional text_as_html field within their metadata for structured reconstruction.
Data Extractor Advanced Features
- Semantic chunking by title to maintain context for AI and RAG applications.
- High-resolution partitioning for accurate table structure inference in complex PDFs.
- Multi-threaded batch processing for high-volume document ingestion.
- Staging functions to easily export data to Pandas DataFrames or JSON strings.
SKILL.md
Loading
Related Openclaw Skills
CV Builder
A professional resume generation skill that converts structured YAML data into polished PDF CVs using multiple industry-standard themes.
Chat with PDF
An advanced AI skill for intelligent conversations, data extraction, and automated summarization of PDF documents.
Batch Processor
Process hundreds of documents simultaneously using parallel execution, progress tracking, and resilient checkpointing logic.
Batch Convert Skill
A unified document conversion pipeline designed to batch process files between formats like PDF, Word, Markdown, and HTML using parallel execution.
Developer Slides (dev-slides)
An AI skill for generating interactive, developer-focused presentations using the Slidev framework and Markdown.
Document Parser (Docling)
An advanced document understanding skill powered by IBM's docling for parsing complex layouts, tables, and figures.
Featured*
