Task-oriented walkthroughs. Each is a complete, copy-pasteable job or scrape call.
| Recipe | What it shows |
|---|---|
| load-more-academic-profile.md | Harvest every publication behind a "Load more" button (the Sidney Fels case). |
| feed-an-llm-with-chunking.md | Hand a large result set to an LLM in digestible chunks (chunk transform + paginated /scrape). |
| paginated-api-harvest.md | Pull a big link-set N-at-a-time with page_size + cursor. |
| change-watching-browser.md | Watch a JS-rendered region for change and only emit what's new. |
| login-session-scraping.md | Log in with fill/press/click, then scrape gated pages. |
| custom-actions-plugin.md | Add a custom recipe step with @register.action. |
| amazon-product-prices.md | Harvest normalized products (title/image/price) with the amazon_search source + webhook ingest. |
Background: ../BROWSER.md (browser layer), ../JOBS.md (jobs/transforms/sinks), ../PLUGINS.md (plugin authoring).