| name | screen-edit |
| description | Load focused context for a single screen-map (parent + children + related + sitemap entries + recent agent log + optional Playwright screenshot). Use when the user runs `/screens edit <id>` or asks to "load context for screen X", "show me ppt/foo", "what's the current state of <screen>". |
Screen-Edit Skill
Pulls everything an agent needs to work on a single screen-map into a markdown summary printed in chat. Avoids hunting for context across files.
When to use
- User invokes
/screens edit <id> (e.g. /screens edit ppt/building-detail).
- User asks "load ", "show me ppt/foo", "what's the state of reality/property-detail".
- Before making any change to a screen-map, to load context.
Inputs
<id> (required) — the screen-map id (<product>/<slug>).
--root <path> (optional; defaults to repo root).
--playwright (optional; capture a screenshot if the local app is running).
Implementation
REPO_ROOT="$(git rev-parse --show-toplevel)"
pnpm -C "$REPO_ROOT/frontend" --filter @ppt/screen-map cli edit "$ID" \
--root "$REPO_ROOT" \
${PLAYWRIGHT:+--playwright}
Output handling
- Print the CLI's markdown summary verbatim into chat.
- Then offer next-step actions:
- "Edit this file?" → opens the screen-map markdown for editing.
- "Run
/screens validate?" → verifies cross-references.
- "Run Playwright?" → if
--playwright was not supplied initially.