learn-taste — скилл для Claude Code: правила дизайна из правок
Скилл learn-taste для Claude Code превращает правки по дизайну в правила с объяснениями. Он разбирает замечания из текущей сессии и предлагает, что сохранить для следующих задач.
MIT 13 700 установок ★ 58 678 обновлён 2026-10-11
npx skills add emilkowalski/skills --skill learn-taste -a claude-codeЧто делает learn-taste
Скилл собирает пользовательские правки по дизайну и формулирует из них правила с объяснением причин. Он ищет явные замечания, изменения файлов, повторные просьбы и отклонённые варианты, затем отделяет общие предпочтения проекта от разовых решений для одного экрана.
Перед записью скилл показывает кандидаты и сохраняет только одобренные правила. Если причина правки не названа, он предлагает версию причины как предположение и просит подтвердить её. Код компонентов он не меняет.
Правила по умолчанию попадают в DESIGN.md в корне репозитория; проверяемые по коду ограничения можно оформить как lint-правило, ограничение токенов или тест. Если агент не читает DESIGN.md, скилл предлагает добавить ссылку на него в CLAUDE.md или AGENTS.md.
Какие задачи решает
- собрать повторяющиеся правки по дизайну в правила проекта
- отделить общие предпочтения от решений для одной страницы
- зафиксировать причину пользовательской правки рядом с правилом
- превратить проверяемое ограничение интерфейса в lint-правило или тест
Примеры запросов агенту
- Разбери правки по дизайну в этой сессии и предложи правила для
DESIGN.md. - Найди повторяющиеся замечания по интерфейсу и отдели общие правила от разовых решений.
- Проверь, какие правки можно закрепить lint-правилом или тестом.
Как установить learn-taste
1Через Skills CLI (Vercel)
npx skills add emilkowalski/skills --skill learn-taste -a claude-code2Вручную: положить файлы
~/.claude/skills/learn-taste/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/learn-taste/
1Через Skills CLI (Vercel)
npx skills add emilkowalski/skills --skill learn-taste -a codex2Вручную: положить файлы
.agents/skills/learn-taste/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add emilkowalski/skills --skill learn-taste -a cursor2Вручную: положить файлы
.cursor/skills/learn-taste/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install emilkowalski/skills skills/learn-taste2Вручную: положить файлы
.github/skills/learn-taste/SKILL.md1Через Skills CLI (Vercel)
npx skills add emilkowalski/skills --skill learn-taste -a gemini-cli2Командой платформы
gemini skills install https://github.com/emilkowalski/skills --path skills/learn-tasteВручную: положить файлы
.devin/skills/learn-taste/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Через Skills CLI (Vercel)
npx skills add emilkowalski/skills --skill learn-taste -a opencodeClaude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы о learn-taste
Как установить learn-taste в Claude Code?
Выполни npx skills add emilkowalski/skills --skill learn-taste -a claude-code. Файл скилла появится по пути ~/.claude/skills/learn-taste/SKILL.md.
Что делает learn-taste?
Он превращает пользовательские исправления интерфейса в предлагаемые правила с объяснениями. Записывает только правила, которые пользователь одобрил, и не меняет код компонентов.
Куда learn-taste записывает правила?
По умолчанию правила добавляются в DESIGN.md в корне репозитория. Проверяемое ограничение можно оформить как lint-правило, ограничение токенов или тест.
Исходный текст
SKILL.md · MIT
---
name: learn-taste
description: Turn a session's design corrections into rules, with their why, in DESIGN.md, a project skill, or a lint check.
disable-model-invocation: true
---
# Learning Taste
Every correction a user makes is taste the agent didn't have. If it isn't written down, the next session makes the same mistake and the user corrects it again. This skill finds those corrections, turns each one into a rule with its reason, and writes it where the next agent will read it.
It does ONE thing: write rules. It never changes component code, and it never adds a rule the user didn't express.
## Hard Rules
1. **Only the user's taste.** Every rule traces back to something the user said, edited, or confirmed. Never add a rule because it seems like good practice. That's your taste, and your taste is what the user was correcting.
2. **Never invent a why.** If the user gave no reason, propose one marked *inferred* and get it confirmed before writing. A wrong why is worse than none: the next agent extends the reasoning to cases it doesn't fit.
3. **Propose, then stop.** Present every candidate and write only what the user approves. A wrong rule repeats in every future session, exactly like the mistake would have.
4. **A check beats a sentence.** If a violation can be found by matching code (a banned value, a raw hex instead of a token, a forbidden property), propose a lint rule, a token restriction, or a test. A check runs every time. A sentence in a long file gets skipped when the agent's context is full.
5. **Write for an agent that wasn't here.** No "like we discussed" or "the card from earlier". Name the class of component, the value, the token, the file.
## Workflow
### Phase 1: Collect corrections
Read the current conversation from the first turn. If the user names earlier sessions and you can read their transcripts (Claude Code stores them as JSONL under `~/.claude/projects/<project>/`), read those too.
Explicit corrections are the minority. Look for all five signals:
| Signal | Example | What it tells you |
| --- | --- | --- |
| Explicit correction | "No, semi-transparent shadow, not a border" | The rule, often with the why |
| Silent edit | User changed `duration: 400` to `200` in a file you wrote | Your default was wrong and their value is the rule. Find these through notices that a file changed outside your edits, or by diffing a file against what you last wrote |
| Repeated nudge | "Smaller." → "Still too big." → "Ok" | Your default was off by the total distance. Record where they landed, not the steps |
| Revert or rejection | "Undo that", picking variant B over A, deleting something you added | What not to do. The why is usually missing, so it goes to Phase 3's questions |
| Confirmed non-obvious choice | "Yes, keep the icon outside the button" | A choice a fresh agent might not repeat. Record it only if a fresh agent would plausibly do otherwise |
For each one, capture the user's exact words (or the diff), what you had done, what they wanted instead, and any reason they gave.
Skip what isn't taste: functional bugs, content specific to one screen ("the headline should say Pricing"), and corrections to your understanding of the task.
**Completion criterion:** every turn has been read, and every correction is listed with its source quote or diff.
### Phase 2: Filter
Walk every correction through this tree:
```
Would a fresh agent building a DIFFERENT component in this repo make the same mistake?
├── No (a one-off: this campaign's color, this page's layout) → Skipped
└── Yes
├── Does an existing rule cover it? Check DESIGN.md, CLAUDE.md / AGENTS.md, project skills.
│ ├── Yes, and it says the same thing → The rule failed. Rewrite it stricter
│ │ (a number, a token, "never") instead of adding a duplicate.
│ ├── Yes, and it contradicts → Conflict. Show both; the user picks.
│ └── No → continue
├── Can a violation be found by matching code? → Check candidate.
│ Keep a one-line rule too, so the why is recorded somewhere.
└── Did the user say it applies everywhere, not just here? → Ask where their global rules live.
Otherwise it's a project rule.
```
**Completion criterion:** every Phase 1 correction ends as Skipped, Rewrite, Conflict, Check candidate, or a rule to generalize.
### Phase 3: Generalize and find the why
Pick the narrowest level that would have prevented the mistake on the next, different component:
| Level | Example | Verdict |
| --- | --- | --- |
| Instance | "The pricing card gets a shadow, not a border" | Too narrow: it only protects this card |
| Class | "Raised surfaces (cards, popovers, menus) separate from the background with `shadow-raised`, never a solid border" | Write this |
| Universal | "Never use borders" | Too broad: breaks inputs, dividers, tables |
If the same correction came up on two or more different components, that's evidence for the class level. If two levels both seem right, ask.
Write the rule strictly: values and token names from this codebase, "always" or "never", no "prefer" or "try to". Use softer wording only when the user framed it as a preference.
Then the why. Use the user's reason, in their words, when they gave one. Otherwise write your best candidate, mark it *inferred*, and put it in the questions for the user.
**Self-test before proposing:** read only the rule and its why, then picture the original task. Would an agent that knows nothing else produce what the user wanted? If it could still produce the wrong version, the rule is too vague. Add the value, the token, or the exception.
**Completion criterion:** every kept rule has a level and a why marked as the user's or *inferred*.
### Phase 4: Decide where it goes
```
Check candidate?
├── Yes → lint rule, token restriction, or test, plus a one-line rule in DESIGN.md
└── No
├── Is DESIGN.md's section for this aspect over ~40 lines, AND relevant to only some tasks
│ (motion rules don't matter when building a settings form)?
│ └── Yes → move the whole section into a project skill and leave a one-line pointer in DESIGN.md
└── Default → DESIGN.md at the repo root, under the matching section:
Surfaces · Color · Typography · Spacing & layout · Motion · Interaction · Copy
```
Create DESIGN.md or the section if it's missing.
**Make sure it gets read.** A DESIGN.md that no agent loads changes nothing. If neither CLAUDE.md nor AGENTS.md points to it, propose adding this line to whichever one the project uses: `Before building or changing UI, read DESIGN.md.`
Entry format:
```md
## Surfaces
- **Raised surfaces (cards, popovers, menus) use `shadow-raised`, never a solid border.**
Why: a solid border makes a raised surface look flat and heavy; a semi-transparent shadow separates it from the background and adapts to both themes.
<!-- 2026-11-12: "no, semi-transparent shadow, not a border" -->
```
The comment holds the date and the user's words. When a later correction conflicts, it shows which rule is newer and what the user actually said.
### Phase 5: Report and stop
Present the report below, then stop.
### Phase 6: Write on approval
Write only the approved items. Show the resulting diff of DESIGN.md (and any skill or config file). For a check, implement it and run it across the codebase. If it flags existing code, report the count and ask whether to make it a warning or fix the violations. Never fix them without asking, because that changes component code.
## Required Output Format
### Rules to add
| # | You said | Rule | Why | Where |
| --- | --- | --- | --- | --- |
| 1 | "no, semi-transparent shadow, not a border" | Raised surfaces use `shadow-raised`, never a solid border | Solid borders look flat and heavy; shadows adapt to both themes | DESIGN.md › Surfaces |
| 2 | Changed `duration: 400` → `200` in `Dropdown.tsx` | Dropdowns open in 200ms | *inferred:* 400ms felt sluggish on a surface opened many times a day | DESIGN.md › Motion |
### Needs your answer
Each *inferred* why, each ambiguous generalization level, and each global-or-project question. One line each, with your recommendation.
### Conflicts
The existing rule, the new correction, and which one you'd keep.
### Make it a check
Each check candidate with its exact implementation (the rule config or test), and what it would flag in the current codebase.
### Skipped
One line per skipped correction and why, so the user can override.
Close with: "Say `write all` or `write 1, 3`, and answer the questions above."
If the session had no taste corrections, say so in one line. That's a good result, not a failure.
## Invocation Variants
| Invocation | Behavior |
| --- | --- |
| `/learn-taste` | Current session, full workflow, stop at the report |
| `/learn-taste <sessions>` | Also read the named or most recent N session transcripts |
| `write all` / `write 1, 3` | Write the approved items from the last report |
| `/learn-taste prune` | Read DESIGN.md and propose merges for duplicates, fixes for contradictions, and stricter wording for vague rules. Same propose-then-stop rule |
© emilkowalski · MIT · emilkowalski/skills