All skills
simota avatar

/vision

@c805268
by shingo imotasimota/agent-skills85 stars
15

Directing UI/UX creative work — redesigns, new designs, trend application, Design System construction, Muse/Palette/Flow/Forge orchestration. Use for design direction. Offers a co-design pair mode.

Use this Skill: https://skilld.dev/gh/simota/agent-skills/vision

This session only. Nothing lands on disk.

referencedesign-methodology.md

≈1.2k tokens on demand. Your agent reads this file only when SKILL.md points to it.

Design Methodology

Purpose: Use this file when you need the full process for REDESIGN, NEW_PRODUCT, REVIEW, or TREND_APPLICATION.

Contents:

  • Mode-specific process steps
  • Shared phase model

Mode Processes

REDESIGN

  • Visual audit of the current state
  • Composition audit: evaluate first viewport, hero contract compliance, layout restraint, and page structure against reference/composition-principles.md
  • Competitive and trend analysis
  • Define principles
  • Create 3+ directions
  • Detail the selected direction
  • Define style guide and token direction
  • Prioritize components
  • Prepare delegation plan

Output: direction doc plus component specifications

NEW_PRODUCT

  • integrate research and personas
  • create moodboards
  • define color, typography, spacing, and motion foundations
  • composition audit: apply first-viewport rule, hero contract, and page structure from reference/composition-principles.md
  • draft wireframes with composition principles enforced
  • define token architecture
  • instruct prototype work

Output: design-system foundation plus wireframes

REVIEW

  • run Nielsen's 10 heuristics
  • audit visual consistency
  • run trend-gap analysis
  • run accessibility checks
  • prioritize fixes
  • assign next agents

Output: improvement report plus action items

TREND_APPLICATION

  • select applicable trends
  • check brand alignment
  • propose phased rollout
  • pick pilot targets
  • recommend testing approach when uncertainty is material

Output: trend application plan plus before/after concepts

Shared Phases

Phase Key question
UNDERSTAND what business, user, brand, and technical constraints shape the design?
ENVISION what 3+ directions could solve the problem?
SYSTEMATIZE what tokens, components, states, and breakpoints define the system?
DELEGATE which agents should execute what, in which order?
VALIDATE what evidence shows the direction is coherent, usable, and safe?

Per-Recipe Behavior Notes

Referenced from SKILL.md -> Subcommand Dispatch.

  • direction: 3+ options + trade-offs; always attach business-outcome metrics (task-success / time-on-task / conversion lift) to each option.
  • redesign: In REDESIGN mode, modernize the current state while preserving brand consistency. Ask First if scope is 3+ pages. Always reference the brand subcommand result if present.
  • trend: TREND_APPLICATION mode — limit to 2026 trends (AI-driven UI / Calm UI / Adaptive Systems / DTCG v2025.10); changes that break product identity are forbidden. Present a before/after concept diagram.
  • system: Superset of NEW_PRODUCT. Always produce a distribution plan to Muse/Palette/Flow/Forge. Make the Core → Brand → Product token hierarchy explicit.
  • brand: Vision strategy + brand alignment. Always define primary palette / typography pair / 5 voice keywords / 5 anti-keywords. Apply orchestrated inheritance for multi-brand. Always read the Compete report if present.
  • moodboard: Pre-ENVISION stage. For each of 3-5 directional axes, gather reference images / palette / fonts / tone keywords; narrow 9 candidates → 3 finalists. List the differentiation axis and risk per finalist.
  • audit: REVIEW mode. Output Nielsen 10 heuristics / WCAG 2.2 AA contrast & focus & target-size as pass/fail; detect token drift and prioritize the remediation backlog (P1/P2/P3) by effort × impact.
  • multi: Tri-engine design-direction generation — parallel Codex/Antigravity/Claude subagents (each 2–3 directions from a loose prompt), Concurrence-Divergence scoring, aesthetic-spectrum coverage, GROUND checks (brand-fit / persona-fit / WCAG 2.2 AA / reference-existence / outcome-link / AI disclosure), and mandatory downstream handoff stubs. Portfolio merge is default; use multi --compete only when the user explicitly wants a single direction. Details → Multi-Engine Mode below and reference/tri-engine-direction.md.
  • pair: Interactive co-design (INTERACTIVE — the dialogue is the deliverable). Vision drives, user navigates; propose 2-3 options → choose → produce/delegate → confirm one decision at a time. Runs within any Operating Mode; distinct from multi (one-shot portfolio) and direction (batch doc). VERIFY: one decision at a time (no full-direction dump), every option carrying rationale + trade-offs + a measurable outcome metric + WCAG 2.2 AA note (never "looks better"); Vision wrote no code (artifacts come from delegated cluster skills — Muse/Forge/Flow/Palette/Frame/Prose); per-decision user-choice + confirmation gate before production (never auto-lock, even under AUTORUN — draft the plan, return Next: USER); bounded by max-decisions (12) / user-stop / direction-locked / diminishing-returns; VALIDATE (dark-pattern / WCAG / handoff-readiness) at close. Full contract → reference/co-design-pair.md.

Source: SKILL.md on GitHub

No alerts14d5 checks · Risk SAFE
  • Gen Agent Trust Hub14d

    The 'vision' skill is a comprehensive design-direction tool that operates by defining strategies and delegating work to other specialized agents. It contains robust internal guardrails, including mandatory accessibility standards (WCAG 2.2 AA) and ethical AI principles. No security risks were identified.

  • Socket14d

    No alerts

  • Snyk14d

    Risk: LOW · No issues

  • Runlayer6mo

    9 files scanned · No issues

  • ZeroLeaks5mo

    Score: 93/100 · 2 sections analyzed

Signed by skilld at c805268. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 days ago.

Activeupdated 2 weeks ago

README badge

README badge for simota/agent-skills/vision