Jev 101
Start Guides
Search
中文 EN
Media heat 1 · Quiet★ 0heat incomplete (no X)

Sunagaku Brand Check

A MIT local app that keeps human brand criteria separate from two independent evaluators—OpenAI Decisions and hosted TypeSafe Jev (TYPESAFE_API_KEY)—and reproduces synthetic comparison runs into reports/latest.json.

Repo: sean-sunagaku/sunagaku-brand-check ↗

What it is

Compare brand criteria without mixing the human rubric into the model prompt soup.

What it does: Side-by-side OpenAI Decisions vs hosted Jev brand checks.

Why it belongs in the showcase: Transparent dual-evaluator brand gate with reproducible eval.

Caveats: JP docs; synthetic cases.

Evidence

Public MIT repo, created 2026-11-03, 0 stars as of 2026-11-03. README (Japanese) documents loopback demo, TYPESAFE_API_KEY / OPENAI_API_KEY, eval harness with 6 synthetic cases × 2 evaluators × instruction A/B, and design.md.

Caveats

  • Docs/UI primarily Japanese
  • Synthetic samples; not a general brand-accuracy claim
  • 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.