fix(canvas): show full Agent prompt tooltips

Merged
BI
2mo ago
+10-2

Summary

Separates compact canvas display values from tooltip labels so long Agent prompts are no longer shortened before the tooltip receives them. Agent cards keep the existing 50-character first-message preview, block sizing, and auto-layout behavior, while editor and read-only preview tooltips receive the complete first-message content. Long plain strings such as Function code remain unchanged.

The renderer now supports an optional full tooltip value and opens the tooltip when the visible text is clipped or when it was shortened upstream. Password masking, selector hydration, regular row tooltip behavior, and structured-value summaries are unchanged.

Type of Change

  • Bug fix
  • New feature
  • Breaking change
  • Documentation
  • Other

Testing

  • bun run test lib/workflows/subblocks/display.test.ts in apps/sim: 28 tests passed
  • bun run test src/workflow-block/sub-block-row-view.test.tsx in packages/workflow-renderer: 1 test passed
  • bun run type-check in apps/sim
  • bun run type-check in packages/workflow-renderer
  • Scoped Biome check across all 7 changed files
  • git diff --check

Reviewers should focus on the display-versus-tooltip separation: the compact Agent chip must remain unchanged while its tooltip exposes the complete first message, including when the compact string itself fits and was truncated upstream.

Checklist

  • Code follows project style guidelines
  • Self-reviewed my changes
  • Tests added/updated and passing
  • No new warnings introduced
  • I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

Screenshots/Videos

Before behavior and both comparison screenshots are in the linked Slack thread. No after screenshot is attached because the compact preview and layout intentionally remain visually unchanged; only the tooltip content changes, and that behavior is covered by the focused renderer regression test.