rust-error-handling-review

Request an error handling review scoped to your staged or unstaged changes, a branch comparison, or named files. The review traces errors from origin to final handler and reports findings with severity, location, and concrete fixes without editing code.

It detects discarded results like `.ok()` and empty error arms, lost context in `.map_err`, panic paths from `unwrap` and indexing, missing retries or timeouts, cleanup failures, and unobserved spawned task outcomes. Each finding requires a reachable failure condition and consequence.

Both are valid choices and the review does not mandate one library per layer. It only reports when a real caller loses required recovery information or must depend on hidden representations, regardless of which error library is used.

No, four or more `#[from]` variants do not establish overuse. The review traces what each conversion preserves and what callers need, reporting context loss only when a relevant operation or recovery distinction is actually lost.

Do not run it as a repository-wide audit when the diff is empty; the review is scoped to the requested target. It also does not change dependencies or error strategy, and reports findings without editing unless fixes are explicitly requested.