Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Extracts structured data and clean markdown from websites using the Tabstack API. Supports JSON schema-based extraction with pre-built templates for jobs, news articles, product pages, and contact info. Includes retry logic, 24-hour caching, and batch processing for multiple URLs.
Schema-based extraction returns consistent JSON across many pages without writing custom per-site parsers.
When to use it
- Scraping job listings from company career pages
- Pulling product details from e-commerce sites
- Extracting article text as clean markdown for summarization
- Batch processing a list of URLs into structured JSON
- Archiving news articles with consistent metadata fields
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Scrape job listings from https://careers.example.com and return structured data
- 1Select or customize job_schema.json from references/ directory
- 2Run extraction: bb scripts/tabstack.clj json https://careers.example.com references/job_schema.json
- 3Parse returned JSON for title, company, location, salary, and requirements fields
- 4Retry on failure using json-retry command if initial request times out
JSON object with job title, company name, location, salary range, requirements array, and apply URL
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.