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

Jev Prompt Reversi

A MIT Reversi lab that feeds board features as state and a natural-language POLICY to hosted TypeSafe Jev (jev-1.14.0 via the TypeSafe SDK) to pick the next move, with a code clone player for comparison.

What it is

Prompt a decision model to play Reversi, then compare with a code twin.

What it does: Computes board features, asks hosted Jev which legal move fits the POLICY, and logs matches in a desktop GUI.

Why it belongs in the showcase: Clear TypeSafe SDK Choice usage on structured game state.

Caveats: Toy domain; verify against current SDK docs.

Evidence

Public MIT repo, created 2026-10-29, 0 stars as of 2026-10-29. README (Japanese) describes JevPlayer using TypeSafe SDK model jev-1.14.0, JevClonePlayer heuristic twin, MonteCarlo/Random baselines, Tkinter GUI, and strategy notes in reversi_strategy.md. HTTP 200 on GitHub.

Caveats

  • Educational game lab, not a production agent product
  • Requires a TypeSafe API key for live JevPlayer games
  • README and strategy notes are primarily Japanese

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.