Feature: Reprocess a selected time range with a different app/prompt (without re-consuming minutes for audio already captured)

[omi-discord-vector]

User pain point (Discord):

• Users want to re-run processing on an already-recorded conversation for a specific time range using a different app/prompt.
• Today, minutes are consumed once audio is processed, and there is no “reprocess segment with different prompt” flow.

Proposed feature:

• Allow selecting a time range within a recording/conversation and re-running a chosen app/prompt against the existing audio/transcript.
• Ideally: no additional cloud “minutes” charged when re-running against already-uploaded audio (or clearly show what will be charged).

UX ideas:

• Conversation → ⋯ → Reprocess → choose range + choose apps/prompts → run
• Status: queued/running/done; ability to compare outputs