Safvr brand design guardrails
Skill Sheshiyer/skill-clusters/skills/safvr-brand-design-guardrails
Apply and protect SAFVR Landingpage2.0 visual system with scaffolding, consistency checks, and brand guardrails. Use when asked to style/refactor SAFVR UI, adjust Tailwind tokens, tune typography or spacing, polish CTA blocks, or review branding drift.From its SKILL.md
npx -y skills add Sheshiyer/skill-clusters --skill safvr-brand-design-guardrailsAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
5.8 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it
SAFVR Brand Design Guardrails
A SAFVR-specific design skill for scaffolding, consistency enforcement, and anti-drift QA across Landingpage2.0.
Target project root:
safvr/Landingpage2.0
What this skill does
- Loads brand context before any visual change
- Applies a deterministic SAFVR design scaffold (tokens, typography, spacing, component shape)
- Runs repeatable consistency checks (colors, fonts, CTA styles, link constants)
- Produces concise brand QA output for every significant UI update
Trigger Signals (Keyword-Aware)
Treat this skill as active when user intent includes terms like:
- Brand work: “safvr branding”, “design dna”, “visual identity”, “brand consistency”
- UI polish: “make it cleaner”, “landing page polish”, “premium look”, “blueprint style”
- System work: “tailwind tokens”, “typography hierarchy”, “spacing rhythm”, “cta style”
- QA work: “consistency check”, “off-brand review”, “design audit”, “visual regression”
If request is SAFVR UI-related and ambiguous, default to this skill.
Hard Gate: Read Before Designing
Before making any visual or UX changes, read in this order:
AGENTS.mdDocs/design-dna.jsontailwind.config.tssrc/app/globals.csssrc/lib/links.ts- Relevant page/component files being changed
Conflict precedence:
- explicit user request
- AGENTS.md constraints
- design-dna.json/tailwind tokens
- existing implementation conventions
No styling work starts before this context is loaded.
Brand Foundation (Non-Negotiable)
Typography
- Display/headings:
font-display(Archivo) - Body/UI:
font-body(Figtree) - No ad-hoc font families unless explicitly approved
Palette + Tokens
- Use
blueprint-*colors from Tailwind config - Prefer semantic tokens/classes over raw hex
- Avoid adding one-off color values if token exists
Visual Language
- Light blueprint aesthetic with crisp linework
- Rectangular CTAs (avoid pill shapes)
- Lucide line icons with restrained stroke
- Pattern/grid effects are supportive texture, not visual noise
- Keep critical copy as HTML (not baked into images)
Spacing + Rhythm
- Preserve generous section spacing
- Reuse nearby spacing cadence; avoid per-section custom scales
Scaffold Workflow (Default for non-trivial design tasks)
Refinement mode (required)
Use a 2–3 pass loop for meaningful SAFVR UI work:
- Pass 1 — Brand alignment: establish token/type/shape truth from references
- Pass 2 — Component normalization: align implementation to canonical patterns
- Pass 3 — Deterministic QA: run harness checks and report findings
1) Brand Snapshot
Capture:
- typography roles used
- token palette for this task
- spacing rhythm and section cadence
- component shapes/states that must remain consistent
- one-line “visual truth” statement
2) Change Plan
For each touched area:
- what changes visually
- which classes/tokens are used
- why it remains on-brand
- responsive impact
3) Implement Minimal Changes
- smallest viable change set
- prefer extending existing patterns over creating parallel systems
4) Consistency Check
Run deterministic checks (see references/prompting.md harness commands).
5) QA Summary
Return:
- files changed
- checks passed/failed
- exceptions with rationale
- optional follow-up opportunities
Deterministic Consistency Checker
Token Drift
- detect newly introduced raw hex colors where blueprint token should be used
- detect hardcoded font stacks bypassing
font-display/font-body - detect accidental “pill” CTA radius patterns on primary actions
Structural Integrity
- CTA links must use centralized constants (
src/lib/links.ts) - typography hierarchy must remain coherent (display vs body roles)
- spacing should align with neighboring section rhythm
UX + Readability
- maintain usable contrast on light blueprint surfaces
- keep motion subtle/purposeful and reduced-motion safe
- prevent key text from being moved into raster assets
Output Contract
For substantial design tasks, always return:
- Brand Snapshot (2–6 bullets)
- Implemented Changes (file-by-file)
- Consistency Findings (pass/fail)
- Exceptions (if any)
- Next Recommendations (optional)
Anti-Patterns
- introducing arbitrary colors when blueprint tokens exist
- skipping AGENTS/design-dna read before styling
- mixing unrelated visual styles in one flow
- moving critical copy into static images
- decorative over-animation
- one-off fixes that increase visual entropy
If exceptions are required, document them clearly.
When to read reference files
| Task | File |
|---|---|
| Full SAFVR token table, typography roles, spacing + visual rules | references/design-system.md |
| Reusable UI patterns: buttons, cards, nav, sections, forms | references/components.md |
| AURA phase semantics, ribbon behavior, and color/motion boundaries | references/aura-visual-language.md |
| AURA implementation adjudication using good-vs-bad examples | references/aura-examples.md |
| Task workflow, consistency/audit commands, keyword triggers, pre-ship checklist | references/prompting.md |
| Tailwind/global CSS hook points and safe extension patterns | references/tailwind-config.md |
What ships with it: 6 files
19.8 KB alongside SKILL.md
references/
- aura-examples.md2.3 KB
- aura-visual-language.md2.9 KB
- components.md4.6 KB
- design-system.md4.2 KB
- prompting.md3.4 KB
- tailwind-config.md2.4 KB