- Added `Run` model to track scraper invocations, including metadata such as status, input, and output.
- Implemented `ToolOutputSpill` model to store context-editing spills separately from user-facing logs.
- Updated middleware to handle spill placeholders and integrate with the new models.
- Enhanced REST API to record runs and expose run history through new endpoints.
- Adjusted tests to validate the new run logging functionality and ensure proper integration with existing capabilities.
web.crawl scrapes a single URL (maxCrawlDepth=0) or spiders a whole site,
backed by the proprietary site_crawler engine. Rewires the scraping subagent
tools and capability tests onto the new verb.
