Topic
Entity intake — medications, supplements, and what touches the system
Entities are external inputs that change state. Capturing them cleanly prevents false confidence in downstream recommendations.
Explainer
Framework note
An Entity is any structured input that changes state: conditions, prescriptions, supplements, peptides, and other categories the system knows how to respect. If the model cannot see it, it cannot honor timing, overlap, or risk.
Why this matters
Incomplete entity data is one of the fastest ways to get confident wrong answers. The reasoning layer will interpolate around holes you left—and you will not see the interpolation, only the headline.
How it works in the model
Entities sit beside Baseline and Body Signals as hard constraints. They are not a moral ledger; they are load and interaction geometry. The app’s flows exist so “what you actually take” matches “what the system assumes,” before pathways or protocols tighten.
What people get wrong
Reframe: Omitting a compound is not privacy—it is silent confounding. The failure mode looks like “the app doesn’t get me” when the gap is missing inventory.
What’s next
Keeping entities current is what makes downstream logic trustworthy. Join the waitlist to use Re:Formd when it launches.
Join the Waitlist
Be first to access the Re:Formd system.
Curious about the human expert layer? Meet Sabrina.