taste-scorecard

Applies a two-pass design-taste review to player-visible features before and after building.

1|Updated Jun 29, 2026
One-click install
npx skills add https://github.com/Raynos/kami-kakushi --skill taste-scorecard-raynos
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: taste-scorecard
Source: https://github.com/Raynos/kami-kakushi/tree/main/.claude/skills/taste-scorecard
Command: npx skills add https://github.com/Raynos/kami-kakushi --skill taste-scorecard-raynos

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Design standards often sit unread in documentation while shipped features drift from them. This Skill operationalizes a 21-principle taste standard by forcing a constraint brief before any player-visible work is built and a scored review after, so design intent and actual results are compared side by side. ## Core Features & Use Cases - Pass 1 constraint brief: Walks all 21 principles and 4 values before building, producing one concrete line per applicable principle that shapes the work. - Pass 2 scorecard: Re-walks all principles after building, scoring each variant with pass/fail/n-a marks and tagging failures as briefed misses or blind spots. - HR-item integration: Attaches the brief and per-variant scorecards to review items so stated intent and scored results are read together. - Use Case: When designing a new game UI surface or narrative beat, generate the taste brief first, build variants against it, then score each variant and file the results for human review. ## Quick Start Run the taste two-pass flow on the new feature I am about to build, starting with the constraint brief.

Frequently Asked Questions about taste-scorecard

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I apply a design taste standard to new features?

Run the two-pass flow: before building, walk all 21 principles and write one concrete constraint line per applicable principle; after building, re-walk all principles and score each variant pass, fail, or not applicable. Attach both artifacts to the review item.

What is the difference between the constraint brief and the scorecard?

The brief is Pass 1, written before building, stating what the work must do to honor each applicable principle. The scorecard is Pass 2, written after, recording whether it actually did. The diff between them reveals execution slips versus blind spots in the standard.

When is the taste review exempt from running?

The flow is exempt for the diverge section-one set: one-line tweaks, copy edits, token nudges, and bug fixes. Everything else that ships player-visible, including features and narrative beats, gets both passes.

What does a blind spot tag mean in a scorecard?

A blind spot tag marks a failed principle that the Pass-1 brief never named, meaning the taste standard's text failed to fire for that case. These tags accumulate in review items as evidence for sharpening that principle later.

Is the taste scorecard an automated verification gate?

No, it is a skill step plus mandatory review-item sections, not a verify gate, because taste is judgment. The enforcement is artifact-shaped: a review item missing its brief or scorecard blocks is visibly incomplete.