Agents heat 1 · Quiet★ 0heat incomplete (no X)

Claude JEV auto-effort

A MIT Claude Code mod that picks per-request effort from plan step, prompt, and stage, using hosted TypeSafe Jev (TYPESAFE_API_KEY → api.typesafe.ai) as judge with scrubbed payloads and word-list fallback.

What it is

Pick effort with a typed judge, keep a hill-climb log of misses.

What it does: Hosted Jev (or word lists) chooses Claude request effort.

Why it belongs in the showcase: Explicit TypeSafe judge with privacy scrubbing and fallback.

Caveats: Claude-specific.

Evidence

Public MIT repo, created 2026-11-03, 0 stars as of 2026-11-03. README documents Auto/Manual band UI, Jev scrubbing (URLs/keys/paths), /effort-mode jev|rules, miss reporting, and Claude desktop zip install.

Caveats

  • Claude desktop plugin zip install path
  • Scrubber is best-effort; unlisted names can leak
  • 0 stars

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.