Skill Claude CodeCodex
Analyze Google Analytics data, review website performance metrics, identify traffic patterns, and suggest data-driven improvements. Use when the user asks about analytics, website metrics, traffic analysis, conversion rates, user behavior, or performance optimization.
Skill Claude CodeCodex
A safety gate for adding reference entries to a catalogue. It checks whether the newer Core v2 catalogue system is ready before allowing new design references.
Skill Claude CodeCodex
Reference batch writer compatibility gate during the DESIGN.md Core v2 catalog cutover.
Skill Claude CodeCodex
Legacy catalog component writer safety gate. Use when component harvesting is requested, but stop before editing until the Core v2 graph-backed catalog writer and readers are available.
Skill Claude CodeCodex
Legacy survey-based DESIGN.md writer redirect. Use for old omd:design requests only to route them to the Core v2 Autopilot or init workflow; this compatibility skill never writes DESIGN.md itself.
Skill Claude CodeCodex
Archived DESIGN.md impact experiment compatibility gate. Use when an old Lab #01 run is requested, but do not generate new legacy 9/12-section documents; direct new experiments to the Core v2 benchmark track.
Skill Claude CodeCodex
Legacy deep-reference writer safety gate. Use for old brand migration requests only to prevent retired section/frontmatter writes until the Core v2 catalog migration pipeline is authoritative.
Skill Claude CodeCodex
Legacy reference token writer compatibility gate during the DESIGN.md Core v2 catalog cutover.
Skill Claude CodeCodex
Compatibility router for older omd invocations. Route existing UI work to omd:apply, autonomous greenfield work to omd:autopilot, and DESIGN.md-only setup or migration to omd:init. Use it only when the legacy skill name is invoked; it reads Core v2 or legacy context without writing DESIGN.md itself.
Agent Claude Code
Stage 0 deterministic gate of the eval pipeline. Validates DESIGN.md Core v2 binding and portability, then runs accessibility, URL, state, and forbidden-copy checks. Pass/fail is mechanical, never opinion-based.
Agent Claude Code
An agent that decides which visual assets a page or component needs and how to provide them. Assets can include icons, illustrations, charts, photos, logos, video, or 3D renders.
Agent Claude Code
Channel-aware image materializer. Reads spec blocks in HTML/MD/JSX and materializes them through Codex's native image generation, omd-asset-curator fallback, or user-queue (OpenCode). One spec format, three downstream paths.
Agent Claude Code
Reads the full run output, user feedback, and persona ABANDONs, then writes a root-cause critique of omd-master's decisions. Forces re-entry at the lowest broken phase rather than surface patches. No write tools beyond critique.md — the constraint is intentional.
Agent Claude Code
Read-only design-system architect for OmD Autopilot. Derives a project-specific semantic system, provenance, component/state coverage, and unresolved decisions from the prompt and repository. Never edits DESIGN.md or product files.
Agent Claude Code
Visual + brand consistency reviewer. Audits HTML/MD/JSX against the brand DESIGN.md across typography, color budget, radius scale, component states, mobile responsiveness, spacing. Outputs BLOCK / WARN / FYI with line refs. Read-only advisory — never modifies artifacts.
Agent Claude Code
Read-only final-gate critic. Enforces an 8-item rubric. Hard 2-round revision cap. Forbids 'looks good' rubber-stamps and requires line refs for every FAIL.
Agent Claude Code
A Korean, English, Japanese, simplified-Chinese, and traditional-Chinese editor for product text and multilingual user interfaces.
Agent Claude Code
Korean blog / long-form writer with 6 voice presets (toss-tech-design, karrot-neighborly, brunch-maker-popular, biz-formal-report, academic-paper, journalism-broadsheet). Default preset toss-tech-design.
Agent Claude Code
An agent that adapts Korean source content into English, Japanese, Simplified Chinese, and Traditional Chinese versions. Adaptation preserves the original meaning, facts, instructions, and URLs while fitting each language’s conventions.
Agent Claude Code
A guided design coordinator that examines a repository, asks only needed questions and leads a design process from user journeys through wireframes, components, copy and checks. It uses checkpoints to control each stage.
Agent Claude Code
Writes all UI text (button labels, error messages, empty states, success confirmations, onboarding copy) strictly conforming to Core v2 Content & Locales. Refuses forbidden phrases. Never invents tone.
Agent Claude Code
Coordinates multi-agent design workflows across writing, five-locale adaptation, humanization, UI slop audit, design review, final QA, and image materialization. Maintains a two-round revision cap and logs every handoff.
Agent Claude Code
Adversarial synthetic user that walks through generated UI under a strict persona prompt with hard turn budget and ABANDON token. Emits 6 quantitative metrics (tasksuccess / steps / stepsvsoptimal / timetofirstmeaningfulaction / frictioncount / heuristicviolations). Never emits SUS / NPS — those are theatre.
Agent Claude Code
A read-mostly auditor that finds repeated interface and copy patterns which make product screens feel generic. It separates design-pattern issues from ordinary quality or accessibility problems and personal preferences.