Code Quality

Code Quality

Operations

• review <pr-id|commit-range|description>: review the identified change. For a PR, follow REVIEW.md; it owns practice routing, evidence, verdicts, and posting. For a commit range, apply the same relevant practices and report findings to the caller without inventing a PR destination.
• cleanup [scope]: implement removal of redundant error handling in the requested scope. The existing defensive-code-cleanup command routes here.

Read code quality and select additional guidance from the documentation index. Searches identify candidates; inspect actual behavior before reporting a violation.

Error-Handling Cleanup

Search the requested source for catches that only log and rethrow, fabricate a generic success/empty result, or duplicate a framework-owned error response. For an unbounded cleanup request, select up to ten related, justified removals.

Before removing each catch, trace its caller and owning error boundary. Establish what the caller observes after removal and why the existing handler is redundant. Do not remove:

• meaningful rollback, resource cleanup, retry, or recovery;
• domain-error conversion to a contractual HTTP response;
• explicitly owned best-effort operations or per-item failure isolation;
• authentication, permission, or other fail-closed security handling;
• required cross-version behavior or expected external-reference misses.

Apply fallback rules, deployment compatibility, and reference authority when those boundaries are involved. A catch containing only a log is not automatically redundant: verify whether it owns necessary diagnostics or rejection handling.

Remove unused imports and adjust callers only when the preserved contract requires it. Keep async work attached to its owning cancellation signal; detach() is not a replacement for error handling inside signals.