Faster code gates: a Rust hook, oxlint, and runner-bound CI
A per-edit hook went from 210ms to 15ms and lint from 21s to 0.7s. The bottleneck was process startup, not code checks, and CI was bound by runners.
A per-edit hook went from 210ms to 15ms and lint from 21s to 0.7s. The bottleneck was process startup, not code checks, and CI was bound by runners.
A failed git push piped into tail still exited 0, and the script moved on. set -o pipefail makes the pipeline report the failure.