Pull request 6 had stale description text and open review feedback spanning indexer handlers, test support, and notification-hook reads.
The branch needed repo docs and GitHub metadata to match the current ERD indexer implementation state before merge.
Decision:
Tighten the reviewed handler paths by normalizing optional string inputs, hardening allocation helpers, removing notification hook list-and-scan reloads, and improving shared test support determinism.
Keep REST search request routes documented as API-key-protected control-plane endpoints while preserving the existing system-actor behavior required by current search-request flows.
Replace the stale PR description with an accurate summary of the shipped indexer scope and reply to each open review comment with the action taken.
Consequences:
Review feedback is resolved with code, test, and GitHub metadata aligned to the current branch state.
The notification hook write path now reloads by primary reference instead of depending on list ordering.
Search request control-plane handlers still rely on the system actor until a future authenticated user-to-actor mapping exists.
Follow-up:
Revisit indexer REST actor attribution if authenticated app users gain stable public-id mapping in the API layer.
Remove any remaining outdated review threads after maintainers confirm the closeout comments.