feat: ground persona reactions in the whole scrolled page and whether it works - #29
Merged
Merged
Conversation
… it works Personas opened the three above-fold viewport shots and grounded reactions in static measurements. Two additions, both from evidence the collector already captures — no extra browser work, still deterministic: - The persona now reads scroll.foldScreenshots, so it reacts to the whole scrolled page instead of the hero alone. "I'd bounce before the CTA" now depends on how far down the CTA actually is. - It grounds reactions in behavioral evidence, not just how the still looks: interactivePromises.results (did the mobile menu open, did the anchor jump), focusPass (keyboard reachability + focus rings), mobile touchTargets, hoverStates. Each persona pulls what its designPriorities care about; a null pass means it didn't run and must not be invented. So "would this work for me" is answered by whether the flow works, not only by how the hero looks.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The scroll-fold + interaction-evidence enrichment. Both additions use evidence the collector already captures — no extra browser work, still fully deterministic. It makes every persona read more grounded, across all page types.
What changed
The persona evaluator used to open the three above-fold viewport shots and ground its reaction in static measurements. Now it also:
scroll.foldScreenshots— the below-fold view, so it reacts to the whole scrolled page instead of the hero alone. "I'd bounce before the CTA" now depends on how far down the CTA actually is, not a guess from the first screen.interactivePromises.results— did the mobile menu open, did the anchor jump, did the accordion expand (passed). "The menu didn't open" is a hard bounce for a rushed-mobile-user.focusPass— keyboard reachability + visible focus rings (withoutIndicator,nonSemanticClickables).viewports.mobile.touchTargets— thumb-sized taps.hoverStates— feedback on hover.Each persona pulls the evidence its
designPrioritiescare about; a null pass means it didn't run and must not be invented.Why this over live per-persona browsing
Discussed and rejected the alternative (each persona driving the browser live): it breaks determinism, the read-only boundary, and the one-collector principle. This gets ~the same behavioral truth from evidence collected once — reproducible, no extra agents.
Testing
Full suite green (1123, 0 fail).
vision-personas.test.jsguards the wiring (reads fold screenshots, consults interactivePromises/focusPass/touchTargets, frames it as whether the page works, never fabricates a missing pass).