WeVue reads your brand book, reviews creative against your rules, and pins exactly what’s wrong — on the asset, with severity and proof. And when a task hits the review step in Optimizely CMP, the verdict lands right back in the task.
Meridian Tea Co.
100% natural · naturally loved
“cure” implies a health benefit — restricted language in marketed claims.
confidence 0.94
Teal #14B8A6 sits outside the declared palette (moss, cream, bark).
confidence 0.88
“100% natural” needs substantiation on file before use.
confidence 0.79
analyzed by gpt-5.6-terra · 14 rules applied · overall confidence 0.87
No new place to check, no batch exports. WeVue reviews where the work already is.
Assign the review step to WeVue in your Optimizely CMP workflow and connect your workspace. Nothing else changes.
When a task reaches the step, WeVue pulls its assets and checks each one against your rules and brand book — findings pinned, severity and confidence attached.
One comment lists every asset’s result with links to the full canvas. The step completes itself when everything is clean — and stays open if anything needs a human.
Attribution is recorded per step — which provider, which model, whose key. Bring your own API key; it’s envelope-encrypted at rest and used only for your reviews.
Mark findings as good catches or false positives. Known false positives are suppressed in later reviews, and any finding can be promoted into the rule library in one click.
“off-palette color usage” — flagged again
Import your brand book and WeVue proposes rules grounded in verbatim quotes, verified against the document server-side. Anything it can’t cite never reaches your library.
Health-benefit wording must match the approved claims list.
source: “…no product communication may imply therapeutic benefit…”
Set up your brand, connect your workflow, and let the first review run today.