Skip to content

DonkeyKing01/tasteful-ui-skill / tasteful-ui

tasteful-ui

Verified

UI design and implementation for real product surfaces with taste-first critique, reference routing, project-specific design briefs, variation comparison, and implementation verification. Use when Codex should redesign, build, polish, or critique frontend UI by reading project context, exploring taste, routing through `references/catalog.md`, synthesizing `PROJECT_DESIGN.md` or `design.md`, using investment gates to avoid wrong taste/reference/scope, implementing in the existing stack, and evaluating whether the result is actually better. Best for app screens, dashboards, landing pages, query tools, product surfaces, and design explorations; not for backend-only work or tiny cosmetic edits.

FileKindSize
SKILL.mdskill md8.0 KB
eval/ui_result_critique.mdmarkdown757 B
formats/PROJECT_DESIGN.template.mdmarkdown2.7 KB
modes/design_critique_only.mdmarkdown474 B
modes/production_ui_implementation.mdmarkdown791 B
modes/taste_first_redesign.mdmarkdown1.1 KB
references/catalog.mdmarkdown9.4 KB
references/designs/ai_llm_platforms/claude.mdmarkdown19.8 KB
references/designs/ai_llm_platforms/cohere.mdmarkdown14.4 KB
references/designs/ai_llm_platforms/elevenlabs.mdmarkdown14.8 KB
references/designs/ai_llm_platforms/minimax.mdmarkdown13.9 KB
references/designs/ai_llm_platforms/mistral.mdmarkdown15.3 KB
references/designs/ai_llm_platforms/ollama.mdmarkdown15.2 KB
references/designs/ai_llm_platforms/opencode.mdmarkdown15.2 KB
references/designs/ai_llm_platforms/replicate.mdmarkdown13.8 KB
references/designs/ai_llm_platforms/runwayml.mdmarkdown14.0 KB
references/designs/ai_llm_platforms/together.mdmarkdown13.9 KB
references/designs/ai_llm_platforms/voltagent.mdmarkdown25.7 KB
references/designs/ai_llm_platforms/xai.mdmarkdown15.9 KB
references/designs/automotive/bmw.mdmarkdown9.6 KB
references/designs/automotive/bugatti.mdmarkdown26.0 KB
references/designs/automotive/ferrari.mdmarkdown24.9 KB
references/designs/automotive/lamborghini.mdmarkdown20.6 KB
references/designs/automotive/renault.mdmarkdown23.7 KB
references/designs/automotive/tesla.mdmarkdown21.7 KB
references/designs/backend, database _ devops/clickhouse.mdmarkdown15.0 KB
references/designs/backend, database _ devops/composio.mdmarkdown20.4 KB
references/designs/backend, database _ devops/hashicorp.mdmarkdown15.3 KB
references/designs/backend, database _ devops/mongodb.mdmarkdown14.7 KB
references/designs/backend, database _ devops/posthog.mdmarkdown18.1 KB
references/designs/backend, database _ devops/sanity.mdmarkdown20.7 KB
references/designs/backend, database _ devops/sentry.mdmarkdown14.4 KB
references/designs/backend, database _ devops/supabase.mdmarkdown14.5 KB
references/designs/design _ creative tools/airtable.mdmarkdown3.5 KB
references/designs/design _ creative tools/clay.mdmarkdown16.9 KB
references/designs/design _ creative tools/figma.mdmarkdown11.3 KB
references/designs/design _ creative tools/framer.mdmarkdown17.2 KB
references/designs/design _ creative tools/miro.mdmarkdown4.7 KB
references/designs/design _ creative tools/webflow.mdmarkdown4.0 KB
references/designs/developer tools _ ides/cursor.mdmarkdown18.6 KB
references/designs/developer tools _ ides/expo.mdmarkdown18.2 KB
references/designs/developer tools _ ides/lovable.mdmarkdown17.0 KB
references/designs/developer tools _ ides/raycast.mdmarkdown18.1 KB
references/designs/developer tools _ ides/superhuman.mdmarkdown18.5 KB
references/designs/developer tools _ ides/vercel.mdmarkdown19.0 KB
references/designs/developer tools _ ides/warp.mdmarkdown16.3 KB
references/designs/e-commerce _ retail/airbnb.mdmarkdown13.7 KB
references/designs/e-commerce _ retail/meta.mdmarkdown20.0 KB
references/designs/e-commerce _ retail/nike.mdmarkdown19.5 KB
references/designs/e-commerce _ retail/shopify.mdmarkdown19.5 KB
references/designs/fintech _ crypto/binance.mdmarkdown18.4 KB
references/designs/fintech _ crypto/coinbase.mdmarkdown4.9 KB
references/designs/fintech _ crypto/kraken.mdmarkdown4.2 KB
references/designs/fintech _ crypto/mastercard.mdmarkdown26.4 KB
references/designs/fintech _ crypto/revolut.mdmarkdown8.4 KB
references/designs/fintech _ crypto/stripe.mdmarkdown20.0 KB
references/designs/fintech _ crypto/wise.mdmarkdown8.3 KB
references/designs/media _ consumer tech/apple.mdmarkdown19.9 KB
references/designs/media _ consumer tech/ibm.mdmarkdown19.1 KB
references/designs/media _ consumer tech/nvidia.mdmarkdown16.7 KB
references/designs/media _ consumer tech/pinterest.mdmarkdown11.6 KB
references/designs/media _ consumer tech/playstation.mdmarkdown27.5 KB
references/designs/media _ consumer tech/spacex.mdmarkdown11.1 KB
references/designs/media _ consumer tech/spotify.mdmarkdown12.7 KB
references/designs/media _ consumer tech/theverge.mdmarkdown27.7 KB
references/designs/media _ consumer tech/uber.mdmarkdown18.2 KB
references/designs/media _ consumer tech/vodafone.mdmarkdown29.4 KB
references/designs/media _ consumer tech/wired.mdmarkdown22.0 KB
references/designs/productivity _ saas/cal.mdmarkdown17.4 KB
references/designs/productivity _ saas/intercom.mdmarkdown6.3 KB
references/designs/productivity _ saas/linear.mdmarkdown21.9 KB
references/designs/productivity _ saas/mintlify.mdmarkdown17.0 KB
references/designs/productivity _ saas/notion.mdmarkdown17.9 KB
references/designs/productivity _ saas/resend.mdmarkdown18.0 KB
references/designs/productivity _ saas/zapier.mdmarkdown18.3 KB
taste/anti_generic_rules.mdmarkdown918 B
taste/taste_critic.mdmarkdown1.7 KB
taste/taste_exploration.mdmarkdown2.6 KB
workflows/implementation.mdmarkdown1.1 KB
workflows/variation_first.mdmarkdown778 B
workflows/verification.mdmarkdown975 B

taste/taste_exploration.md

View on GitHub
# Taste Exploration

Use this before choosing references.
The goal is to discover the right taste direction for this product, not to decorate the UI.

## Provenance

This module is a synthesized operating guide, not a verbatim extraction from one source.
It combines:

- Claude-style design collaboration: explore context, ask useful questions, show taste judgment early, avoid filler.
- Reference-system thinking: convert reusable style traits into concrete choices without copying brands.
- Agent-executable design brief thinking: turn taste into explicit rules an implementation agent can follow.
- Variation-first thinking: compare meaningful directions when taste is uncertain.
- UI critique principles: judge hierarchy, clarity, affordance, visual weight, consistency, and user-task success.

## Taste Questions

Answer these from project context before opening detailed references:

- What is the product really doing for the user?
- What should the surface make the user feel: trust, speed, calm, control, delight, precision, warmth, authority, focus?
- What is already working visually and should not be erased?
- What is weak, generic, noisy, unclear, cramped, or visually dishonest?
- What should become more visible: task entry, data, comparison, status, identity, next action?
- What would make this UI worse even if it looked more designed?

## Direction Axes

Explore along meaningful axes:

- calm vs energetic
- editorial vs operational
- light workspace vs dark cockpit
- sparse search tool vs dense analysis surface
- institutional trust vs consumer warmth
- data-first vs narrative-first
- existing identity preserved vs visual reset
- single-task clarity vs multi-module command center

## Required Taste Readout

Before reference selection, summarize:

- product identity in one sentence
- strongest existing visual cue
- weakest current UI problem
- recommended taste direction
- one direction to avoid
- why this recommendation makes the UI more useful, not just more styled

This readout is the second investment gate.
Ask the user which taste direction is worth investing in, not whether the process should continue and not which reference brand sounds good.

## Taste First, Reference Second

Do not choose a reference because the product category sounds adjacent.
For example:

- EV does not automatically mean dark automotive cockpit.
- AI does not automatically mean gradients or black glass.
- Data dashboard does not automatically mean Linear.
- Search/query tools often need restraint more than visual novelty.

If the taste readout suggests the existing product already has the right emotional tone, use references lightly.
brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).Stars: 323kUpdated Claude CodeCodexVerifiedInstall
brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).Stars: 198kUpdated Claude CodeCodexVerifiedInstall