Cloudflare described its multi-agent security workflow on October 7, explaining how Managed Defense combines collected evidence with specialist AI analysis.

Original factual graphic based on the dated company announcement; capabilities and availability require the stated qualifications.
Original factual graphic based on the dated company announcement; capabilities and availability require the stated qualifications.

The company says fixed application code gathers data and enforces scope before models interpret it. Analysts receive related alerts, evidence gaps and recommended next steps.

Original factual graphic based on the dated company announcement; capabilities and availability require the stated qualifications.
Original factual graphic based on the dated company announcement; capabilities and availability require the stated qualifications.

Cloudflare reports that an earlier broad agent produced unsupported claims. The published design keeps evidence collection separate from interpretation and supports human analyst review.