AGENT SKILL

ui-iteration

Reusable screenshot-driven UI polish workflow for iterative visual refinement. Use when improving layout, hierarchy, spacing, feedback, or Tailwind-based presentation through repeated review-and-adjust passes. DO NOT USE FOR: functional bug fixing, accessibility-only audits, or major redesign strategy work (use Code-Smith or frontend-design).

适用职业设计师
使用场景前端开发
License 许可
MIT
最近更新
2026-08-16
Stars
★ 3
Forks
0

01 前言 Front Matter

name
ui-iteration
description
Reusable screenshot-driven UI polish workflow for iterative visual refinement. Use when improving layout, hierarchy, spacing, feedback, or Tailwind-based presentation through repeated review-and-adjust passes. DO NOT USE FOR: functional bug fixing, accessibility-only audits, or major redesign strategy work (use Code-Smith or frontend-design).

03 SKILL.md 正文

<!-- platform-assumptions: markdown skill guidance for VS Code custom agents in Agent Orchestra; assumes the calling agent owns browser-tool applicability, invocation boundaries, and implementation handoff decisions. --> <!-- markdownlint-disable-file MD041 MD003 -->

UI Iteration

Reusable methodology for screenshot-based UI polish.

When to Use

  • When a page or component feels visually off but needs targeted refinement rather than redesign
  • When repeated screenshot comparison is the fastest way to validate polish work
  • When the work is mostly spacing, hierarchy, theme coherence, or interaction feedback
  • When Tailwind or presentation-layer tweaks should be applied in bounded iterations

Purpose

Improve the UI through short, measurable polish loops. Each pass should identify a small set of visual issues, apply focused changes, and confirm that the after state is visibly better than the before state.

Iterative Polish Workflow

  1. Capture a baseline screenshot of the target page or component state.
  2. Assess the current UI against the evaluation criteria below.
  3. Select 3-5 concrete improvements for the pass.
  4. Implement the smallest set of presentation changes that address those issues.
  5. Capture a follow-up screenshot.
  6. Compare before and after, and record whether the pass produced measurable improvement.
  7. Repeat until the planned iteration count is complete or additional changes stop producing clear gains.

Screenshot Requirements

  • Use a representative populated state rather than empty placeholders
  • Prefer standard desktop browser width unless the task is explicitly mobile-first
  • Keep captures focused on the target surface, not the full desktop
  • Use programmatic screenshots when browser tools are available; otherwise follow the repo's fallback process

Screenshot output: save screenshots to .tmp/ per skills/terminal-hygiene/SKILL.md ## Scratch & Temp-File Hygiene.

Aesthetic Evaluation Criteria

Generic UI Principles

Principle What to Check
Readability Text contrast, font sizes, line height
Visual Hierarchy Clear focal points and emphasis
Spacing Consistent padding, margins, and breathing room
Alignment Grid alignment and edge consistency
Consistency Similar elements styled similarly
Feedback Hover, active, loading, and empty-state clarity

Product-Facing Checks

Criterion What Good Looks Like
Information Clarity Primary data and actions are obvious at a glance
Theme Coherence Colors, typography, and surfaces feel intentionally related
Data Legibility Statuses, values, and labels remain easy to parse
Motion Restraint Motion supports comprehension rather than distracting
Pattern Benchmark New surfaces fit the product's established visual language

Tailwind And UI Polish Heuristics

  • Prefer design tokens and semantic utilities over arbitrary values
  • Use spacing scales consistently before introducing custom exceptions
  • Improve hierarchy with contrast, weight, grouping, and whitespace before adding more chrome
  • Fix misalignment at the container or layout level before patching individual elements
  • Keep motion subtle and purposeful
  • Prefer utility composition over custom CSS when the design system already supports the need

Iteration Parameters

Parameter Default Typical Override Example
Iterations 5 Polish Dashboard 3 times
Scope Full page Polish just the SummaryCard component
Focus All Focus on spacing and alignment

Output Formats

Per-Iteration Analysis

## Iteration N/5 Analysis

**Target**: [Page/Component]

**Assessment**:

- ✅ [What is working]
- ⚠️ [Minor issues]
- ❌ [Significant issues]

**Proposed Improvements**:

1. [Specific change] - [Rationale]
2. [Specific change] - [Rationale]
3. [Specific change] - [Rationale]

**Files to Modify**:

- `src/...`

Final Session Summary

## UI Polish Complete

**Target**: [Page/Component]
**Iterations**: N/N

**Changes Made**:

1. [Change] (Iteration 1)
2. [Change] (Iteration 2)

**Before/After Summary**:

- Spacing: [Before] -> [After]
- Hierarchy: [Before] -> [After]
- Feedback: [Before] -> [After]

**Remaining Suggestions**:

- [Optional future polish]

Related Guidance

  • Load frontend-design when the polish work needs a stronger visual point of view
  • Use browser-tool instructions from the current repo for navigation, startup, and cleanup details

Gotchas

Trigger Gotcha Fix
The before and after screenshots look alike The pass changed code without producing visible improvement Reduce scope and choose 3-5 changes with clearer visual impact
Trigger Gotcha Fix
Tailwind tweaks pile up element by element The layout becomes inconsistent and harder to maintain Fix spacing and alignment at the layout or container level first

04 同仓库更多 Skills

Grimblaz
Grimblaz/agent-orchestra

webapp-testing

Playwright end-to-end testing guidance for web apps, with practical patterns and setup steps. Use when creating or improving browser-based E2E coverage, test stability, and CI execution. DO NOT USE FOR: React component tests (use ui-testing), canvas game interaction (use browser-canvas-testing), TDD workflow (use test-driven-development), debugging failures (use systematic-debugging), or randomized property tests (use property-based-testing).

测试与质量前端开发
★ 3
Grimblaz
Grimblaz/agent-orchestra

verification-before-completion

Evidence-based verification checklist before marking work complete. Use before PRs, releases, marking tickets done, or any "I'm finished" declaration; also when naming a baseline commit for a differential suite claim, when designing a mutation campaign or red-run evidence, when making an absence claim, a universal claim, or a one-time demonstration offered as a standing guarantee, when attributing a measurement or claiming a sweep or verification loop is complete, and when stating what a check or a CI suite will do about a change. DO NOT USE FOR: post-merge cleanup or archival (use post-pr-review) or processing GitHub review comments (use code-review-intake).

★ 3
Grimblaz
Grimblaz/agent-orchestra

validation-methodology

Reusable validation and review methodology for staged validation, failure triage, and prosecution-depth setup. Use when running validation ladders, triaging failures, or executing adversarial review passes. DO NOT USE FOR: CE Gate orchestration, specialist dispatch ownership, or step execution flow (keep those in Code-Conductor).

★ 3
Grimblaz
Grimblaz/agent-orchestra

upstream-onboarding

Shared opening-phase protocol for upstream agents (Experience-Owner, Solution-Designer, Issue-Planner) and Code-Conductor when invoked on an existing GitHub issue. Renders a scaled context brief and runs a standards check on inherited work at each phase boundary. Use when a user-invocable agent receives a request referencing an existing GitHub issue and must load solution-authoring and upstream-onboarding at the opening phase. DO NOT USE FOR: subagent dispatches (which already operate within an assessed session context); post-merge review (use post-pr-review); research or non-tree subagents.

人力资源Skill与智能体开发
★ 3
Grimblaz
Grimblaz/agent-orchestra

ui-testing

Resilient React component testing strategies focusing on user behavior. Use when writing or reviewing component-level React tests, fixing flaky component tests, or establishing React testing patterns. DO NOT USE FOR: Playwright E2E tests (use webapp-testing), canvas game interaction (use browser-canvas-testing), TDD workflow (use test-driven-development), debugging failures (use systematic-debugging), UI component design (use frontend-design), or randomized property tests (use property-based-testing).

设计师测试与质量前端开发
★ 3
Grimblaz
Grimblaz/agent-orchestra

tracking-format

Tracking-file frontmatter and coordination format guidance. Use when creating or reviewing local tracking files and related handoff conventions. DO NOT USE FOR: session-memory plan authoring rules outside the documented references here.

★ 3