Scrapling adds adaptive crawling, AI-safe scraping

This is a strong “agent-era scraping” package: it treats DOM drift, anti-bot friction, and crawl orchestration as core problems instead of edge cases.

• –Adaptive element tracking is the standout feature; it can relocate targets after layout changes, which cuts down on brittle selector maintenance
• –The fetcher stack covers plain HTTP, stealth browsing, and proxy rotation, so teams can choose the least expensive path that still works
• –The spider framework plus pause/resume makes it more than a parser library; it looks like a real crawl orchestration layer
• –MCP and the `--ai-targeted` mode are the tell that Scrapling is aiming at AI workflows, not just classic scraping
• –The MCP path is a good sign, because scraping for agents is quickly becoming a security problem as much as a data problem