mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-15 23:31:29 +00:00
64bd591ccf
Applying multiple AI-suggested templates schedules each next template via setTimeout(handleUseTemplate, 500) with the timer id discarded, so navigating away mid-queue let the chain keep running: it continued creating guardrails through the API and setting state on the unmounted panel. Store the timer in a ref and clear it in an unmount cleanup. The regression test drives the two-template flow and asserts no further guardrail fetches happen after unmount |
||
|---|---|---|
| .. | ||
| litellm-dashboard | ||
| Dockerfile | ||
| nginx.conf | ||