Remote Control mobile app shows session as 'ready for review' while background Agent/Bash jobs are still running

Summary

In the Claude Code mobile remote-control app, the session list status indicator flips to "ready for review" as soon as the assistant returns a turn to the user — even when the session is actively waiting on run_in_background Bash tasks or background Agent tool calls that haven't completed yet.

This is misleading: you pick up the phone expecting the session is idle and needs your input, only to find it's still chugging through background work and not actually waiting for you.

Reproduction

1. From a remote-controlled Claude Code session, dispatch a long-running background job, e.g.:

• Agent(..., run_in_background: true) (sub-agent in background)
• Bash(..., run_in_background: true) (e.g. a multi-minute test run)

1. The assistant turn ends and returns control while the background task is still in flight.
2. Open the mobile app's session list.

Observed: The session is shown with a "ready for review" / idle status.

Expected: The session should reflect that background work is still running (e.g. "busy", "background tasks pending", or surface the in-flight job count). The "ready for review" state should only appear once all background tasks tied to the session have either completed (with their notifications delivered) or been explicitly acknowledged.

Why it matters

• Lots of agentic workflows depend on background sub-agents / long Bash tasks. The mobile UX assumption that "assistant returned a turn ⇒ user attention needed" doesn't hold for these.
• Users get falsely paged ("ready for review") and then context-switch back to find nothing to do.
• Conversely, when the background task actually completes and the model wakes back up, the visual state on the phone often doesn't change in a way that distinguishes that real wake-up from the earlier false "ready" state.

Suggested fix direction

Track per-session whether any run_in_background Bash processes or background Agent calls are still outstanding, and reflect that as a distinct status state in the mobile session list (e.g. "running in background" vs. "awaiting input"). The "ready for review" label should require: no in-flight background tasks AND the assistant turn has actually ended.

Environment

• Mobile remote-control app (iOS Claude Code app)
• Long-running sessions with run_in_background Bash + background Agent dispatches

Related

Adjacent but distinct from existing issues — none of these describe the session-list status indicator specifically:

• #66700 — model fabricates <task-notification> blocks during long background-task waits
• #65424 — /remote-control hangs with false progress assertions
• #63649 — background task notifications should include exit code and output tail