Agent skill
frontend-design-lumina
Design or redesign frontend interfaces in quant-lab with high-quality, production-ready UI/UX while preserving Lumina identity. Use when the user asks for UI polish, visual redesign, page/layout revamp, dashboard/chart UX improvements, interaction quality upgrades, or anti-generic "AI-looking" frontend output.
Install this agent skill to your Project
npx add-skill https://github.com/CheekyCodexConjurer/quant-lab/tree/main/skills/frontend-design-lumina
SKILL.md
Frontend Design (Lumina)
Use this skill to produce distinctive, production-grade UI without drifting from quant-lab visual identity.
Design intent
Before coding, define in 4 bullets:
- Purpose: what user problem this UI solves now.
- Audience context: trader/dev/analyst and primary scenario.
- Tone direction: one explicit direction (ex: "minimal quant-tech", "dense professional terminal", "calm analytical").
- Constraints: performance, accessibility, existing component primitives, and file-size limits.
Then pick one memorable signature detail (layout rhythm, motion moment, or visual hierarchy choice) and keep it coherent across the feature.
Lumina-first guardrails
- Preserve minimalist, symmetric, clean quant-tech style.
- Prefer compact overlays/menus over heavy panels.
- Reuse existing primitives before creating new patterns.
- Keep spacing and hierarchy consistent; avoid decorative noise.
- Use motion to clarify state/intent, never as visual clutter.
Anti-generic UI checklist
Avoid shipping generic AI-looking UI. Before finalizing, verify:
- No random gradient/glow usage without semantic purpose.
- No inconsistent radius/spacing scale between nearby components.
- No mixed interaction quality (all controls must cover hover/focus/disabled/loading/error as applicable).
- No broken density balance (too sparse or too crowded for the workflow).
- No accidental style drift from neighboring Lumina surfaces.
Implementation workflow
- Start with a small UI brief (purpose, audience, tone, constraints).
- Define component boundaries (view/hook/service/types) to stay LLM-friendly.
- Implement minimal diffs first; add polish after behavior is correct.
- Validate interaction completeness:
- keyboard/focus-visible
- escape/click-outside when relevant
- reduced-motion safe behavior
- viewport fit/flip for floating surfaces
- Run verification with
skills/verify_changes/SKILL.mdand UI smoke when UI behavior changed.
Deliverable standard
In final output, always include:
- chosen visual direction (1 line)
- what was improved in UX/clarity (short bullets)
- risks/trade-offs (if any)
- modified/created files
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
action_logger
Keep an audit trail of changes, commands, and verification.
size_guard
Enforce line budgets and rotate or split large context files.
analyze_repo_capabilities
Detect repo tooling and patterns, then record a dynamic manifest.
safety_validation
Validate changes against .agentignore before commit.
safe_refactor_migration
Plan and execute large refactors with shadow mode and rollback.
command_guard
Enforce .agentpolicy rules before executing commands.
Didn't find tool you were looking for?