Guardrails heat 1 · Quiet★ 5heat incomplete (no X)

jev-engineering

Measured coding-agent tools on Jev (review-router + jev-gate) with public eval kits.

What it is

Public toolkit of coding-agent helpers on TypeSafe Jev: review-router routes PRs to full vs quick review from the diff, and jev-gate is a strict second lock on every command.

What it does: GitHub Action review routing plus Claude/Codex/Cursor/OpenCode hooks; observe mode before enforcement; linked eval kits you can rerun.

Why it belongs in the showcase: Guardrail + staged-review patterns with measured public results, not just README claims.

Caveats: Kit numbers are author-reported; not a blanket security guarantee; payloads leave the machine to the hosted model.

Evidence

Public MIT README + https://eugeniughelbur.github.io/jev-engineering/; review-router and jev-gate with linked public benchmarks.

Caveats

  • Author-reported CVE/injection kit numbers — not a blanket security guarantee
  • Hosted model sees gated payloads / diff context
  • jev-gate is a second lock beside host auto-mode, not a full replacement

Unofficial case page. Stars / list position observed for heat; do not treat heat as a quality SLA. Review the upstream README before you depend on it.