PR 19 review timeout stability

• Status: Accepted
• Date: 2026-04-12
• Context:
  • PR 19 still had unresolved review feedback on a torrent-label test that waited only one second for an emitted settings event.
  • That timeout is short enough to become flaky on contended CI runners even when the event bus behavior is correct.
• Decision:
  • Increase the async event wait in crates/revaer-api/src/http/handlers/torrents/labels.rs from one second to five seconds.
  • Keep the test structure otherwise unchanged because the event subscription contract is still the behavior under test.
• Consequences:
  • Positive outcomes:
    • The test is less sensitive to scheduler jitter and runner contention.
    • The fix is narrowly scoped to the flaky wait boundary instead of changing production event behavior.
  • Risks or trade-offs:
    • A genuine regression in event delivery could take a few seconds longer to fail.
• Follow-up:
  • Implementation tasks:
    • Keep similar async event-listener tests on this branch reviewed for overly aggressive wall-clock assumptions.
  • Review checkpoints:
    • Re-run the repo validation gates and reply on the outstanding PR review threads.

Task Record

• Motivation:
  • The user asked to address all remaining PR feedback, and the only still-actionable comment requested a more CI-stable event timeout.
• Design notes:
  • The change follows the reviewer’s recommendation directly and preserves the current event-stream assertion.
  • The already-open openapi.rs thread was also rechecked locally; the branch already uses std::env::temp_dir().join(OPENAPI_FILENAME), so that thread only needed a fresh reply.
• Test coverage summary:
  • cargo test -p revaer-api update_label_catalog_persists_changes_and_emits_event
  • just ci
  • just ui-e2e
• Observability updates:
  • None. No runtime logging, tracing, or metrics changed.
• Status-doc validation:
  • No README or operator-facing docs needed updates because the change is limited to test stability and ADR/task tracking.
• Risk & rollback plan:
  • Risk is low and limited to test-runtime duration.
  • Rollback is a straightforward revert if the longer timeout proves unnecessary.
• Dependency rationale:
  • No new dependencies were added.
• Stale-policy check:
  • Reviewed:
    • AGENTS.md
    • .github/instructions/rust.instructions.md
    • docs/adr/template.md
  • Drift found:
    • None in policy text; the remaining issue was test timing sensitivity in an existing async assertion.
  • Contradictions removed:
    • None.