Diagnose taste-level alignment and conflicts in creative work.
npx skills add https://github.com/jwynia/the-kepler-testimonies --skill taste-eval
Please help me install this Agent Skill. Skill: taste-eval Source: https://github.com/jwynia/the-kepler-testimonies/tree/main/.claude/skills/taste-eval Command: npx skills add https://github.com/jwynia/the-kepler-testimonies --skill taste-eval
Taste-eval addresses the challenge of aligning creative output with explicit project preferences, surfacing where work diverges and enabling productive creative friction.
Start by running a taste-eval pass on a draft to surface alignment with explicit project tastes.
High-intent search queries and answers about installing and using this skill.
Taste alignment checks evaluate text against stated taste dimensions to surface where drafts align with or diverge from explicit project preferences. The process diagnoses conflicts and provides actionable recommendations to balance creative work.
Diagnosing taste-level conflicts in collaborative storytelling involves identifying where contributors' explicit preferences clash and proposing balancing options. This surfaces productive creative friction by exposing alignment or divergence from defined project tastes during reviews.
Generate diverse voting options by curating choices that reflect different explicit tastes within a group. This process balances contributor preferences and surfaces alignment or divergence, enabling communities to vote on options matching defined taste dimensions.
Yes, taste evaluation applies to drafting, reviewing, and voting scenarios where explicit preferences must be surfaced and balanced. It diagnoses taste-level alignment and conflicts across these stages, exposing divergence and offering actionable recommendations.
Avoid taste alignment checks when a project requires strict adherence to rigid standards rather than balancing subjective preferences. The approach is designed to spark productive creative friction by exposing alignment or divergence from explicit tastes, not enforcing absolute rules.
yu13130122297Identify uncertainties in research framing, datasets, metrics, and baselines.
dayvidphamAutomate PROPOSAL-N plan reviews with ACCEPT/REVISE voting and Beads task updates.
JrakruScaffold Python projects with src, tests, docs, and internal governance separation.
labs21-devAudits AGENTS.md files against four domain principles for implicit gaps.
