Changelog

Alerts have arrived for Custom Intel Agents. Plus a round of improvements across reports, newsletters, and the mobile apps.

Set up alerts to stay on top of your Custom Intel Agent updates

Threat Intelligence

You can now set up email or Slack notifications for any Custom Intel Agent, get notified when your Intel Agent has new updates and configure alerts per AI Agent based on your monitoring needs.

This release brings a few new ways to work with AI across the platform, along with a round of fixes that smooth out crashes, sync issues, and layout bugs. Custom Intel Agents now support Ask AI Analyze and Research actions, IP Insight Cards get a Censys lookup, and Report Builder generation is clearer with a new progress indicator. The rest of the update includes fixes for the mobile app crashes on iPad and iPhone, the Trends Dashboards tab, and a handful of popover and navigation glitches are all resolved.

This release brings upgrades to Ask AI, including a new model tuned for threat actor campaign detection and SPL hunting queries to complement our existing KQL support. We also improved MITRE ATT&CK tagging accuracy and fixed several Ask AI issues around article selection and CVE data retrieval. Alongside these AI improvements, we shipped a batch of reliability and UX fixes across insight cards, the Vulnerability Agent, and table sorting to keep your workflow running smoothly.

CVE Insights Cards now surface exploit type, confidence, and evidence directly on the card, Org Profiles can be @ mentioned across Ask AI, Report Builder, Newsletter, and Custom Intel Agent, and Report Builder citations let you verify LLM claims down to the source passage. Plus a set of Agent Table controls and fixes based on your feedback.

View exploit analysis on CVE Insights Cards

Threat Intelligence

TI analysts can now see the type, confidence, reason, and supporting evidence for every exploit on a CVE Insights Card. This makes it possible to assess at a glance whether a CVE has a real, weaponized exploit, a proof of concept, a detection rule, or no real exploit at all, without leaving Feedly to verify on GitHub or other sources.

@mention Org Profiles across Feedly

Threat Intelligence

You can now define an Org Profile in text (direct input or .md upload) and reference it anywhere in Feedly with an @ mention, including Ask AI (Analyze & Research), Report Builder, Newsletter, and Custom Intel Agent. You can also see everywhere a given Org Profile is referenced, so it's easy to track how it's being used across your workflows.