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 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.
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.