fix(mobile): Normalize attachment and pull-request context ids#80

What Changed

Mobile attachment and pull-request summary records use the shared kind-scoped composer context IDs that web uses, instead of building unchecked IDs. IDs already stored in drafts, queued messages and the outbox are kept when file chips are restored.

Why

Context IDs allow only letters, digits, _ and -, so mobile records built from raw attachment or pull-request identifiers could fail schema validation. The thread-mention half of this fix landed upstream as pingdotgg/t3code#15841 mobile-thread-mention-id; this layer keeps the remaining attachment and pull-request producers consistent with it.

Checklist

• This PR is small and focused
• I explained what changed and why
• I included before/after screenshots for any UI changes
• I included a video for animation/interaction changes