Rank On Maps · Command Brief

Good morning / here's what matters

Built overnight. One real incident caught and contained, the full site-health sweep, and your day ranked. Read top to bottom, it takes two minutes.
Generated
30 Jun 2026
Overnight
5 sites swept
P0 incident · caught + contained overnight

Client sites were occasionally serving the wrong client's website

The overnight sweep caught Emmendorfer's homepage and contact page serving Central Lending Solutions content (wrong business, wrong phone, and a canonical that could make Google merge the roofer into the broker). It flickered between Valor, Skin Avenue and CLS too. It is intermittent edge-level cross-serving, not a build error. As of the last check, every domain is serving correctly again.

Live status: ALL 5 DOMAINS CLEAN · last auto-check passed 12/12 fetches each

Done overnight (safe, no client sites touched)

  • Hard rule saved to memory: a client domain may never serve another client's content, zero tolerance.
  • Continuous monitor installed and running every 30 min, silent when clean, alerts the moment any site shows the wrong brand.
  • Engine cache-hardening added: future site builds mark HTML so it can never be shared-cached across clients.

Staged for your go (needs the Cloudflare layer)

  • Redeploy Emmendorfer + Valor with the new cache headers, then re-verify live. Low risk, but it touches production so I left it for daylight + a QC pass.
  • Ruben handles the Cloudflare-side config (the affected domains' DNS lives outside our CF account). His brief is drafted below, ready to forward.

Do first

ranked
  1. Forward the Cloudflare remediation to RubenP0
    Draft is ready at the bottom of this brief. Gets the cross-serve fixed at the source today.
  2. Skin Avenue: capture the design pick + run the onboarding
    Everything is built and live: design directions · blueprint · E-E-A-T interview.
  3. Two two-minute client wins
    Fix the "veteran-owned owned" typo on Valor's service pages, and add phone-call tracking to Emmendorfer (it's missing). I can ship both on your nod.
  4. Clear the parked HQ decisions
    True MRR source, demo-client cleanup, real WhatConverts account, and the three branch merges. Listed in the blocked-on-you section.

Overnight site-health sweep

5 live sites
SiteVerdictWhat to know
Emmendorfer
roofing · live client
action Homepage + contact were serving CLS (the incident above). Also missing the mandatory phone-call-click tracking, and the non-www apex 404s robots/sitemap before redirect. About + service pages are correct.
Valor
roofing · live client
minor Healthy overall. Tracking, schema, sitemap all good. One copy typo "veteran-owned owned" repeated on two service pages. Caught one flicker of Skin Avenue content (the incident).
CLS
mortgage · live client
healthy Clean across the board. Tracking complete (phone-click wired), 94-URL sitemap, rich schema, AI-crawler allowlist. One harmless one-off cache flap on the sitemap.
Skin Avenue
skin clinic · rebuild target
baseline Current WordPress site has zero analytics (no GA4 at all), Meta Pixel only, no call tracking. 14 pages. This is exactly the gap our rebuild closes. Also flickered CLS once (the incident).
Grays
did not sign
dropped Per your call, removed from the watch list. Not pursuing.

Blocked on you

decisions
  • True MRR source: $28,430 (verified roster) vs $36,331 (Stripe). Pick one, I repoint the whole HQ.
  • Demo-client cleanup: confirm which "active" clients are demo, and whether Austin Area Roofers is a real paying client.
  • Real WhatConverts account: so leads stop coming from the sample account.
  • Three branch merges to hq.rankonmaps.io: backlink-outreach, hq-integrity-fixes, client-brain-sites.
  • Skin Avenue design + Claude Design brand: ROM green or the client's taupe/cream?

Ready to use

links

The guardrail and its log live at ~/projects/rom/ops/. To pause the monitor: launchctl unload ~/Library/LaunchAgents/com.rankonmaps.client-isolation.plist

Forward to Ruben

ready to send
Slack draft · #rom-ops or DM Ruben
Morning Ruben. Overnight monitoring caught a real one: a few of our client sites were intermittently serving the WRONG client's website at the Cloudflare edge. Emmendorfer's homepage + contact were showing Central Lending Solutions (wrong business, wrong phone, and a canonical pointing at CLS, which risks Google merging the two). It also flickered between Valor and Skin Avenue. It's intermittent edge cache cross-serving across Pages projects, not a build bug. Right now all domains are serving correctly again. What I need from you on the Cloudflare side: 1. Confirm each Pages project has only its own custom domain attached (I verified the attachments look correct, so this is likely an edge-cache key issue, not a misattachment). 2. Purge cache on the affected projects/zones: emmendorfer-cd, valor-roofing, cls-rebuild (and the external zones for emmendorferexteriors.com + centrallendingsolutions.com.au, which live outside our CF account). 3. Make sure HTML isn't being "Cache Everything"-cached without a host-aware cache key. I've added a _headers file to the roofing engine that marks HTML as private/no-store, so once we redeploy Emmendorfer + Valor it's hardened. Want a second set of eyes before we push those. I've already: saved a zero-tolerance rule, installed a monitor that checks all client domains every 30 min and alerts on any cross-serve, and added the cache fix to the engine. Happy to pair on the redeploys whenever.