Desktop app hides assistant reply text when the turn includes a ScheduleWakeup call - only the 'Used ScheduleWakeup' chip renders

Description

When Claude responds to a user prompt and calls the ScheduleWakeup tool in the same turn, the desktop app frequently renders only the collapsed tool-use chip ("Used ScheduleWakeup >") and none of the assistant's text for that turn. The text exists in the transcript (the model sees its own prior replies normally on the next turn), but the user never sees it in the app.

Environment

• Claude Code desktop app on Windows 11 Pro (build 10.0.26200)
• Model: claude-fable-5
• Interactive session using self-paced wakeups (ScheduleWakeup, dynamic-loop style) to monitor background/external state between user messages

Steps to reproduce

1. Have Claude monitoring something that makes it arm ScheduleWakeup at the end of each turn (e.g. watching a slow external process).
2. Send any prompt.
3. Claude replies with normal text plus a ScheduleWakeup call in the same turn.

Expected

The assistant's text renders, with the tool call shown as a chip.

Actual

Only "Used ScheduleWakeup >" renders, followed by the working indicator; no assistant text appears. The reporter's screenshot shows the user's prompt, the chip, and nothing else.

Text placement relative to the tool call does not reliably avoid it: turns ordered text-then-call and turns ordered call-then-text both dropped the text. A couple of turns did render text placed after the call, so the failure may be timing-dependent rather than strictly positional.

Observed repeatedly on 2026-08-08 within one session, and the user reports the same behavior in other sessions.

Impact

Any session that uses ScheduleWakeup for self-paced monitoring becomes very hard to hold a conversation in: the user cannot see the assistant's replies and has to ask it to repeat itself, while the assistant has no indication anything went unseen.