菜单
Use when fixing a real visual bug, layout glitch, rendering artifact, or UI regression in arrayview. Requires visual evidence, but keeps the default path targeted unless the bug has broad cross-mode risk.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when implementing any server-side, startup, display-opening, or environment-detection feature in arrayview. Ensures the feature works correctly across all six ways arrayview can be launched — CLI, Python script, Jupyter, Julia, VS Code tunnel, and SSH.
Use when making styling or layout changes to `_viewer.html`. Keeps UI work aligned with arrayview's established visual language without dragging in release-only or cross-mode audit steps by default.
Use when a visual feature touches canvas rendering, zoom, eggs, colorbars, shortcuts, or layout across modes. Keeps implementation and verification consistent without escalating to a full audit by default.
Use when the user explicitly requests a full visual audit, when validating UI work for a release, or when diagnosing a cross-mode visual regression in arrayview. Proactively identifies all affected mode combinations, prescribes per-mode behavior, checks for UI clashes, and runs a Playwright-based visual audit after implementation. Absorbs and replaces the modes-consistency skill.
Use when working through TODO items, implementing batches of features or fixes, or when the user gives multiple tasks at once. Enforces commit-per-item, collateral updates, and cross-mode verification.
Use when the user explicitly asks to sync UI docs/help/test coverage, or when preparing a UI change for release. Ensures visual_smoke.py, help overlay, and docs/ stay in sync.
| name | visual-bug-fixing |
| description | Use when fixing a real visual bug, layout glitch, rendering artifact, or UI regression in arrayview. Requires visual evidence, but keeps the default path targeted unless the bug has broad cross-mode risk. |
Fix visual bugs with evidence, not guesswork. Capture the broken state, fix the cause, capture the result, then run only as much regression coverage as the risk justifies.
tests/ui_audit.py failsDo not use this for brand-new feature design or docs/test sync work.
For a small bug in one known area:
Escalate to the broader audit path only when the bug touches shared layout, shared chrome, zoom, or multiple modes.
| Area | Owner |
|---|---|
| UI structure, CSS, JS layout | _viewer.html |
| Rendering routes | _server.py + _routes_rendering.py |
| Render/compositing helpers | _render.py, _overlays.py |
| Colorbar behavior | ColorBar class in _viewer.html |
Do not treat _app.py as the implementation surface; it is a compat shim.
Use this when the bug has shared-chrome or cross-mode risk:
uv run python tests/ui_audit.py --tier 1
Go to tier 2 only when the affected feature or regression area actually needs it.
_app.py was edited for new behaviortests/ui_audit.py --tier 1 run if the bug touched shared layout/chrome/cross-mode behavior