Skip to content

Lycan

This release improves Cisco Catalyst Center support, extends flow execution timeouts to one week, refines the UI. The word ‘lycan’ comes from the Greek word ‘lykos’ meaning wolf, and lycans in folklore are said to retain their human intelligence even when transformed, unlike traditional werewolves who lose control during the full moon

Improved support for Cisco Catalyst Center 🔐

Section titled “Improved support for Cisco Catalyst Center 🔐”

Ref: #10758

A custom Cisco Catalyst Center credential has been added to improve usability.

Improved support cisco catalyst center

Flow execution timeout increased to one week ⏰

Section titled “Flow execution timeout increased to one week ⏰”

Ref: #10654

Flows can now run for up to one week before timing out, giving you more flexibility for long-running processes and complex workflows that require extended execution time.

Flow execution timeout increased one week

Ref: #11218

We refined the user interface by clarifying callout text, enhancing tooltips for clearer guidance, and updating input modals for a smoother experience. These small but meaningful changes make the platform more intuitive and easier to navigate.

We’ve also improved Studio responsiveness with smarter reuse of recent data (more to come 🙂).

  • Missing borders when browser is zoomed out (e.g., at 90% zoom level) (#10410)
  • Updating credential mapping could send the wrong update for the step (#11060)
  • Pin/unpin chat was hidden in AI··Rooms (#11136)
  • Credential row blinks when updated (#11189)
  • Incorrect application/json default from Catalyst Center parameter schemas (#11194)
  • Unable to pick data input of a FlowHook (#11277)
  • Logs from attached flows are not displayed in AI··Task steps (#11278)
  • Issues during LiteLLM/Azure AI Foundry credential update and creation (#11200)
  • Pills with arrays are not evaluating in HTTP request body (#11211)
  • AI··Rooms with simple agents are missing the first user message (#11233)
  • Unable to open AI··Rooms under certain circonstances (#11285)
  • First call in rooms with simple Agent fails (#11292)
  • An invalid flow can still be deployed successfully (#8060)