mirror of
https://github.com/unclecode/crawl4ai.git
synced 2026-06-10 07:48:50 +00:00
- Route deep-crawl stream=True requests with a single URL through AsyncWebCrawler.arun so each discovered page is streamed as its own CrawlResult - Preserve existing arun_many + MemoryAdaptiveDispatcher behavior for non–deep-crawl or multi-URL streaming. - Add Docker REST tests for deep-crawl streaming success (single URL) and helpful error on multi-URL usage