component-review

Audit one or many existing canonical components in content/components/<id>.yaml against every schema section, the depth contract, recent schema additions, vsRelated bidirectional integrity, vocabulary drift, and source-claim staleness — emit a structured gap report with pre-formatted backlog items ready to drop into docs/backlog.md. Use whenever the user asks to review, audit, check, lint, or QA a component or the whole canon — phrases like "review tabs", "audit modal", "check accordion", "QA-Pass über alle komponenten", "find gaps in X", "atelier-v3 review", "ist X vollständig", "missing was bei Y". This is the systematic per-component version of the broad Atelier-v2 review pass that produced 22 backlog items in one go; running it on stale components catches schema-additions they didn't get migrated for.

Safety Notice

This listing is imported from SkillsMP metadata and should be treated as untrusted until upstream source review is completed.

Copy this and send it to your AI assistant to learn

Install skill "component-review" with this command: npx skills add DominikPieper/skillsmp-dominikpieper-dominikpieper-component-review

No markdown body

This source entry does not include full markdown content beyond metadata.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Open in GitHub

Related Skills

Related by shared tags or category signals.

General

adr-author

Author a new ADR for the uianatomy project under docs/adr/ following the established template (Status / Date / Supersedes / Context / Decision / Rationale / Consequences). Use whenever the user wants to record an architectural decision, deprecate a prior ADR, or document why a proposal was rejected — phrases like "ADR für X", "neue ADR", "decision record", "supersede ADR-NNN", "reject this proposal as ADR", "schreib eine ADR". Picks the next sequential number, links the supersedes target, and sets the status correctly. Prefer this over freehand writing whenever the topic is an architectural choice that shapes the schema, repo layout, MCP tools, or content conventions.

Repository SourceNeeds Review

-0DominikPieper

Automation

backlog-tick

Maintain docs/backlog.md as the single source of workflow truth — flip [ ] → [x] when an item finishes, append the completion date and a one-line outcome plus affected files, update the "Empfohlener Pfad" line if the next item shifted, and append newly discovered work in the right priority bucket. Use whenever the user says "fertig", "done", "erledigt", "gerade gepusht", "ship that", or whenever any non-trivial work concludes; also use mid-session when new follow-ups, bugs, missing tests, deferred cleanups, or schema gaps surface — append them automatically per docs/CLAUDE.md without waiting to be asked. Keeps PX-NN ids sequential within their bucket and prevents stale items from accumulating.

Repository SourceNeeds Review

-0DominikPieper

Research

canon-component-author

Research and author a new canonical component YAML in content/components/<id>.yaml from scratch — synthesizing anatomy, axes, mismatches, mistakes, frameworkMap, motion, responsive, and all P2 sections from W3C APG, MDN, WCAG, and at least three mature headless libraries (Radix, React Aria, Headless UI, Spectrum) plus two design systems (Polaris, Carbon, Material 3, Atlassian, GOV.UK). Use whenever the user asks to add, draft, or research a new canonical component — phrases like "neue komponente X", "draft canonical Y", "research and write the Z anatomy", "add component", "author canon for component", "synthesize component anatomy". Strictly canon-first per ADR-001 — never references implementations like Atelier UI. Validates against the depth contract in shared/tests/depth.test.ts before declaring done.

Repository SourceNeeds Review

-0DominikPieper

Automation

mcp-tool-add

Add a new tool to the uianatomy MCP server (mcp-server/src/server.ts) — Zod input schema, handler that reads from getComponents/getPatterns/getImplementations state, deterministic output shape, error-path for unknown ids, and a matching test in mcp-server/tests/. Use whenever the user wants a new MCP tool, a bulk-fetch wrapper, an aggregation across canon, or a new query surface for AI agents — phrases like "neues MCP-tool", "add tool to MCP server", "expose X via MCP", "bulk-fetch tool", "aggregation tool", "new tool X". Picks the right naming bucket (list_/get_/search_/validate_), writes consistent error returns, and updates the tool-count in docs/backlog.md and project_state memory.

Repository SourceNeeds Review

-0DominikPieper