L03-13a — Select, copy, and capture text and images in Quick #232

Outcome

Operators can select text and tables, copy text, step by line, capture a region, and extract an image from the canvas.

Parent and boundary

Parent: #231. Owner: LoopEditor. This issue owns one bounded slice of Quick Workspace; consume accepted upstream contracts and keep PDF/verdict/publication authority with Core.

Grounding

• No handler: actionSelectText, actionSelectTextAll, actionDeselectText, actionSelectTable, actionCopyText, actionGoToNextLine, actionGoToPreviousLine, actionScreenshot, actionExtractImage.
• Select/Hand tool semantics exist since Select and Hand toolbar buttons are checkable placeholders with no behavior #103; text selection is a separate interaction kind.

Acceptance

• Text selection is a typed interaction state bound to the current revision; a revision change clears it.
• Copied text matches Core's text layer for the selection, including reading order on multi-column fixtures.
• Screenshot and image extraction write through the atomic writer and refuse silently overwriting a file.
• Failure case: selection on a page with no text layer reports that state; it never copies an empty string as success.

Required proof

Focused interaction-controller and quick-canvas tests on text, table, and image fixtures; accessibility smoke for keyboard selection. Record the exact source SHA, test command or hosted run, fixture/artifact identity, and any skipped or unavailable lane. Run the repository's mapped proof and check-change gate for a PR.

Handoff

Record changed files, observed result, remaining risk, and a 1–3 sentence anti-slop review. Completion means acceptance evidence, not merely a merged implementation.

Sources

Approved master roadmap · Gap review of docs/loop-shell-actions.json against Quick command handlers, 2026-10-06.