Meg detail route
Integration Guide
Use this page when the missing piece is when Meg should enter a UI-facing change, what makes a design-review request legible, and what happens after the design verdict lands.
Validated Β· Sep 8, 2026The Meg proof still matches the district contract, the route-shaped request split, and the sibling-route exits that keep Rules, Review, and Scout visible before the lower detail sections begin.
Start here on phone
Keep the handoff answer here, then jump sideways once the route needs the design rule, the review verdict, or the delivery owner instead.
When Meg enters
Decide whether the route needs design direction now or audit after ship
Most readers do not need a command recipe. They need the quick split between pre-build direction, post-ship screenshot review, and the cases where Meg is no longer the right lane.
Use the design gate only when it changes the path
Meg should make the next move clearer. If the question is no longer visual quality, Atlas should route you away instead of adding a fake review hop.
Question
The route is still deciding its visual language or a new UI surface could drift without direction.
Start with
Bring Meg in before build.
Why
Meg is most valuable while the cost of changing course is still low. Early design direction keeps the route from inventing a private grammar that later has to be torn back out.
Examples: new route shells, major visual components, or a UI slice whose hierarchy and component precedent are still uncertain.
Question
The route is built and you need a visible-quality verdict before treating it as ready.
Start with
Bring Meg in after ship for screenshot review.
Why
Rendered UI is the design gate's real proof surface. Meg can evaluate hierarchy, clarity, and drift once the page exists as a screen instead of a plan.
This should end in a concrete punch list or an explicit pass, not in generic reassurance.
Question
The issue is now code behavior, delivery ownership, or governance rather than visual quality.
Start with
Keep moving with Scout or leave for Atticus.
Why
A Meg handoff is useful only when it narrows a visual-design question. Stretching it into a catch-all review step makes the next move less clear, not more.
Use Scout for build and runtime behavior. Use Atticus when the unresolved question is approval-sensitive scope or authority.
Question
Start with
Why
The route is still deciding its visual language or a new UI surface could drift without direction.
Bring Meg in before build.
Meg is most valuable while the cost of changing course is still low. Early design direction keeps the route from inventing a private grammar that later has to be torn back out.
Examples: new route shells, major visual components, or a UI slice whose hierarchy and component precedent are still uncertain.
The route is built and you need a visible-quality verdict before treating it as ready.
Bring Meg in after ship for screenshot review.
Rendered UI is the design gate's real proof surface. Meg can evaluate hierarchy, clarity, and drift once the page exists as a screen instead of a plan.
This should end in a concrete punch list or an explicit pass, not in generic reassurance.
The issue is now code behavior, delivery ownership, or governance rather than visual quality.
Keep moving with Scout or leave for Atticus.
A Meg handoff is useful only when it narrows a visual-design question. Stretching it into a catch-all review step makes the next move less clear, not more.
Use Scout for build and runtime behavior. Use Atticus when the unresolved question is approval-sensitive scope or authority.
What to bring
A good Meg handoff is route-shaped, not tool-shaped
Atlas stays educational here. The useful part is the proof shape Meg needs, not the exact command syntax a particular builder happens to use to gather it.
Name the route and the state you want reviewed
Meg needs to know which page or screen matters and what state the reader should be looking at before she can judge the design clearly.
Good shape: page, intended state, and why that state matters. Avoid asking for a generic vibe check on an undefined surface.
State the visual question, not only the implementation status
A strong request names the design uncertainty: hierarchy, density, component precedent, readability, or whether the route still matches the project language.
βWhat should change?β is weaker than βIs the scan-first hierarchy still clear after this new control rail?β
Bring the relevant design rule or proof pointer
The handoff is stronger when Meg can compare the route against the intended language instead of inferring the design rules from the output alone.
That may be a PDLP pointer, an existing route precedent, or a clear reference to the part of Atlas grammar the builder meant to reuse.
Ask for the next move, not just an opinion
The most useful result is a revised-or-approved next step the builder can act on immediately.
A good Meg review changes the route: tighten spacing, simplify hierarchy, reuse a proven pattern, or proceed as-is.
After review
Meg's verdict changes the next move, not the delivery owner
Meg can sharpen the design rules, return a punch list, or pass the route. The builder still owns the implementation and the follow-through that turns the verdict into a shipped page.