kuat-agent-rules

Create EE-branded slides, web UI, marketing content, graphics, or copy using kuat-agent-docs rules. Use when building or generating new Equal Experts branded artifacts. Always resolves up-to-date rules before creating.

Review EE-branded work against kuat-agent-docs rules. Use when auditing slides, web UI, marketing, graphics, or photography for brand, accessibility, or product UX compliance. Always resolves up-to-date rules before reviewing.

Use when a request needs new visual design concepts or mockups — screens, components, or layouts — before any code or Figma work starts. Determines (1) whether to produce a single design or multiple concept directions, (2) the fidelity level and how strictly Kuat design-system compliance applies at that fidelity, and (3) whether the output route is Figma or Claude Design (inline mockup). Hands off to figma-build-intake or mockup-build-intake accordingly. Not for code implementation (see create-web-app) or slides/imagery (see create-presentation, create-imagery).

Build or modify Equal Experts web application UI — forms, dashboards, app screens, navigation, settings, tables, product flows. Use when creating product UI (React/Vue/HTML) that should follow EE brand, design tokens, accessibility, and the Kuat component system. Not for marketing pages (use web-marketing reference) or slides.

Run before generating or updating any Figma design, screen, or component (before figma-generate-design or use_figma). Confirms whether the work is Equal Experts/Kuat or a different client's own design system, and produces a short build plan. Deliberately separate from Figma's own skills — it does not customise figma-generate-design, figma-use, or any other Figma-provided skill, so Figma's own updates to those skills are never affected.

Run before building a design mockup with the Claude Design inline-mockup tool (show_widget / visualize). Confirms design-system context (EE/Kuat vs. a different client), resolves the fidelity level's compliance bar, and — for Kuat work — pulls real token values and component visual specs live from kuat-agent-rules rather than generic Claude Design tokens, so the mockup is as spec-accurate as the fidelity level requires instead of only directionally similar.

Use when a request needs new visual design concepts or mockups — screens, components, or layouts — before any code or Figma work starts. Determines (1) whether to produce a single design or multiple concept directions, (2) the fidelity level and how strictly Kuat design-system compliance applies at that fidelity, and (3) whether the output route is Figma or Claude Design (inline mockup). Hands off to figma-build-intake or mockup-build-intake accordingly. Not for code implementation (see create-web-app) or slides/imagery (see create-presentation, create-imagery).

Build or modify Equal Experts web application UI — forms, dashboards, app screens, navigation, settings, tables, product flows. Use when creating product UI (React/Vue/HTML) that should follow EE brand, design tokens, accessibility, and the Kuat component system. Not for marketing pages (use web-marketing reference) or slides.

Run before generating or updating any Figma design, screen, or component (before figma-generate-design or use_figma). Confirms whether the work is Equal Experts/Kuat or a different client's own design system, and produces a short build plan. Deliberately separate from Figma's own skills — it does not customise figma-generate-design, figma-use, or any other Figma-provided skill, so Figma's own updates to those skills are never affected.

Run before building a design mockup with the Claude Design inline-mockup tool (show_widget / visualize). Confirms design-system context (EE/Kuat vs. a different client), resolves the fidelity level's compliance bar, and — for Kuat work — pulls real token values and component visual specs live from kuat-agent-rules rather than generic Claude Design tokens, so the mockup is as spec-accurate as the fidelity level requires instead of only directionally similar.

Review Equal Experts web application UI for brand, accessibility, and product-UX compliance — audit a form, dashboard, app screen, or product flow against EE rules. Use when checking existing product UI (code, Figma, screenshots, URLs). Produces findings, not a redesign. Not for marketing pages or slides.

Create or select Equal Experts visual assets — icons, infographics, illustrations, diagrams, and photography selection/briefs. Use when producing or choosing imagery that must follow EE brand, colour, and photography rules. Includes a light pre-publish quality/accessibility check. Not for full slide decks or web UI.

Create Equal Experts branded slide decks and presentations — pitch decks, knowledge-sharing talks, case studies, reports. Builds from the genuine EE master template (inheriting real layouts, the left bracket, the official logo, and Lexend) — never improvises a lookalike. Use when building or editing EE slides. Not for web UI or standalone marketing pages.

Review an Equal Experts slide deck or presentation for brand and content compliance — audit slides (deck file, PDF export, screenshots, or Figma) against EE layout, typography, photography, and voice rules, AND against the genuine brand assets. Use when checking existing decks. Produces cited findings, not a redesign.

Create or select Equal Experts visual assets — icons, infographics, illustrations, diagrams, and photography selection/briefs. Use when producing or choosing imagery that must follow EE brand, colour, and photography rules. Includes a light pre-publish quality/accessibility check. Not for full slide decks or web UI.

Create Equal Experts branded slide decks and presentations — pitch decks, knowledge-sharing talks, case studies, reports. Builds from the genuine EE master template (inheriting real layouts, the left bracket, the official logo, and Lexend) — never improvises a lookalike. Use when building or editing EE slides. Not for web UI or standalone marketing pages.

Review an Equal Experts slide deck or presentation for brand and content compliance — audit slides (deck file, PDF export, screenshots, or Figma) against EE layout, typography, photography, and voice rules, AND against the genuine brand assets. Use when checking existing decks. Produces cited findings, not a redesign.

Review Equal Experts web application UI for brand, accessibility, and product-UX compliance — audit a form, dashboard, app screen, or product flow against EE rules. Use when checking existing product UI (code, Figma, screenshots, URLs). Produces findings, not a redesign. Not for marketing pages or slides.

Add a brand asset (logo variant, photograph, or icon) to the studio asset pack and register it in assets.manifest.json. Use when extending assets/slides/ — e.g. a dark/colour logo lockup, the first EE photography, or a catalogued icon — so create-presentation can insert the genuine asset and review-presentation can compare against it. Contributor-only; repo-local. NOT for recreating/redrawing a logo (never do that) and NOT for building a deck.

Add or edit a guideline or pattern in the passive reference library (reference/). Use when writing/changing brand, design-language, content, accessibility, or per-medium reference — keeps it passive (the WHAT, no procedure), places media patterns under media-types/<medium>/patterns/, updates MIGRATION-MAP.md when paths move, and routes colour changes through the token SoT. Contributor-only; repo-local. NOT for writing a skill (that's author-skill) or applying the brand to a deliverable (consumer skills).

Scaffold a new skill in the Equal Experts house style. Use when adding a new activity skill (consumer-facing, packaged) or a new contributor skill (repo-local) — produces a SKILL.md with a sharp trigger, progressive disclosure into reference/, the link conventions the build expects, a delivery checklist, and a version stamp. Wires it into the plugin build when it's a distributable skill. Contributor-only; repo-local. NOT for editing reference (that's author-reference).

Label and prune the PRESENTATION / SLIDE-DECK master's slide layouts in assets/slides/assets.manifest.json. Slides-only. Use when the slide master ships more layouts than are catalogued (today 65 layouts, only 4 labelled) — enumerate them, label each by purpose, mark which create-presentation may use, and prune unused layouts to shrink the master. Contributor-only; repo-local. NOT for building a deck (that's create-presentation), and NOT for web/marketing/imagery.

Change Equal Experts brand COLOURS at the source and regenerate the downstream artifacts. Use when adding/editing a brand colour, scale step, or semantic alias — anything in reference/design-language/tokens/colors.tokens.json. Edits colours HERE (the SoT), regenerates the human-readable colours.md, runs the drift check, and hands off the kuat-core variables.css half. Contributor-only; repo-local. NOT for applying colours in a deck/app (that's the consumer skills).

Re-slim and brand-fix the Equal Experts PRESENTATION / SLIDE-DECK master template (the .pptx behind kuat-studio) when the raw Google Slides master changes. Slides-only. Use when a new "2026 EE branded master slides" export needs turning into the shipped assets/slides/ee-master-2026.pptx — strips example slides + orphan media, sets the theme font to Lexend, embeds Lora + JetBrains Mono, removes Montserrat, and self-checks. Contributor-only; repo-local. NOT for building a deck (that's create-presentation), and NOT for web/marketing/imagery.

Review the passive reference library — either gate a change before it merges (pass/fail) or do a general review/audit of specific reference file(s). Use when reviewing an edit/add/move under reference/ (a guideline, pattern, token, or colours change) OR auditing an existing reference doc — runs the passive test (no verbs/checklists/role cards/loading tables), link integrity, structure conformance, and token drift, then a manual read, and reports findings by severity. The gate the recent out-of-phase content edit lacked. Contributor-only; repo-local.

Create EE-branded slides, web UI, marketing content, graphics, or copy using kuat-agent-docs rules. Use when building or generating new Equal Experts branded artifacts. Always resolves up-to-date rules before creating.

Review EE-branded work against kuat-agent-docs rules. Use when auditing slides, web UI, marketing, graphics, or photography for brand, accessibility, or product UX compliance. Always resolves up-to-date rules before reviewing.