Document best-effort run-store cleanup error semantics

Co-authored-by: Eric Allam <eric@trigger.dev>

Author: cursoragent

docs/tasks/streams.mdx

@@ -622,6 +622,9 @@ The default payload sent to your task is a rich, typed object that includes:
 optional `runId`, and the underlying `error` (non-Error throws are normalized to `Error`
 instances).
 
+Run-store cleanup is handled as best effort, and cleanup failures won't mask the original
+transport failure that triggered `onError`.
+
 ```ts
 import type { TriggerChatRunState, TriggerChatRunStore } from "@trigger.dev/ai";
 

packages/ai/CHANGELOG.md

@@ -15,3 +15,4 @@
 - Added optional `onError` callback support for observing non-fatal transport issues.
 - Added phase-aware `onError` reporting across send, stream-subscribe, reconnect, and stream-consumption paths.
 - Added normalization of non-Error throw values into Error instances before `onError` reporting.
+- Added best-effort run-store cleanup so cleanup failures do not mask root transport errors.

packages/ai/README.md

@@ -140,6 +140,9 @@ The callback receives:
 - `runId` (may be `undefined` before a run is created)
 - `error`
 
+Cleanup operations against custom `runStore` implementations are best-effort. If store cleanup
+fails, the original transport error is still preserved and surfaced.
+
 ## Reconnect semantics
 
 - `reconnectToStream({ chatId })` resumes only while a stream is still active.