Scraping skills
Preloaded by /scrape
| Skill | Purpose |
|---|---|
scrapling | Default framework (Scrapling 0.4.14) when the agent writes scraper code. |
scrape-operator | Orchestrates jobs: Scrapling first, then scrape (Rust/httpx), then browser. Built for this module. |
web-extractor | Firecrawl-style readable extraction into Markdown/JSON for RAG or migrations. |
data-quality-agent | Deduping, schema consistency, missing-field checks on exported datasets. |
playwright-browser | Drive Chromium when static HTML is an empty shell. |
deep-web-research | Multi-source research that often seeds a scrape list. |
report-generator | Executive summaries next to the raw exports. |
spreadsheet-analyst | Pivot, filter and sanity-check CSV/XLSX outputs. |
Complementary skills
| Skill | Purpose |
|---|---|
prompt-injection-defender | Treat untrusted page content as data, not instructions. |
rag-knowledge-builder | Index a cleaned scrape corpus into a knowledge base. |
fact-checker | Verify claims extracted from pages before reuse. |
competitor-intelligence | Structured competitor monitoring that can feed crawl seeds. |
seo-technical-auditor | Crawlability / structured-data follow-ups after a site scrape. |
Skills load automatically with /scrape; invoke any of them explicitly for a focused task ("use data-quality-agent on scrape/out.csv").