The public Marketing Center page at https://sail-wiki.onrender.com/marketing-center.html is not actually pulling fresh Litify numbers.
The page fetches https://sail-marketing-api.onrender.com/api/data. That endpoint is online, but it serves a deployed static snapshot with as_of = 2026-05-22. Its backend loads services/marketing-api/data_snapshot.json; it has no Salesforce client, no SOQL refresh code, and no scheduled refresh job.
Render check: sail-marketing-api (srv-d89q955ckfvc738t24q0) has only ANTHROPIC_API_KEY and PYTHON_VERSION configured. It is missing the SF_* variables needed for Litify.
A separate service, sail-marketing-dashboard (srv-d87snvjbc2fs73ebak5g), does have Salesforce credentials and is refreshing. Its /data/marketing.json payload was generated on 2026-05-27T07:15:48-04:00.
The wiki page is wired to the wrong refresh lane. It uses sail-marketing-api, which is a static snapshot/chat API. The real daily Litify refresh currently lives in sail-marketing-dashboard.
Litify/Salesforce is already working on sail-marketing-dashboard but not on sail-marketing-api.
Local vault keys are present for CallRail, Google Ads, RingCentral, Semrush, Notion, Pipedream/Zapier, Google service account, Slack, GitHub, Anthropic, OpenAI, and Render. These can support connector syncs, but the current public API does not use them.
Make sail-marketing-api the durable public API and connect it to the refreshed source. Either add Salesforce credentials and live snapshot generation there, or have it transform the fresh sail-marketing-dashboard payload into the flat kpis shape expected by the wiki page.
Add a daily Render cron for the API, add health fields for data_as_of, snapshot_age_hours, and litify_configured, and make the front-end show a stale-data warning when the source is more than one day old.
C:\Users\SAguiar\Documents\Codex\salesforce-litify\marketing_center_live_data_diagnosis_20260527_112441\DIAGNOSIS.md
No production data was changed.