Goal match
Does this plant’s primary indication line up with what you asked for? Weighted highest because nothing else matters if the plant isn’t for the job.
Every recommendation is weighed by published evidence, your constraints, and the plant’s fit for the chosen goal. Each match comes with an explanation and a confidence band — never a single number, never a popularity rank.
— Every plant gets scored on the same five axes. The composite is a weighted blend — never a single black-box number.
Does this plant’s primary indication line up with what you asked for? Weighted highest because nothing else matters if the plant isn’t for the job.
Meta-analyses and RCTs outrank case reports. Quality scores are inherited from the source paper, not invented locally.
Contraindications from EMA/HMPC and the rest of the literature. A plant is downgraded if it conflicts with your constraints.
Form fit (tea / tincture / capsule), preparation time, and shelf availability. A great molecule you can’t source doesn’t help.
A small bonus for under-prescribed plants with solid evidence — so the same five candidates don’t always win.
— Pick a goal, set your constraints. The list re-ranks on every change — composite scores, full reasoning, and a direct link to the plant profile.
— The engine is not a diagnosis tool. Every score is auditable. Every weight is documented. Here’s the short version.
Search volume and citation count are deliberately ignored. A plant that’s common in blog posts but thin on trials will not outrank a less-known plant with strong RCTs.
Tick “pregnancy-safe” and contraindicated plants drop off the list entirely. We don’t bury them — we exclude them. The same goes for medication conflicts.
The engine code is open in the repo. Run the same query twice with the same plants and you get the same number. No model, no temperature, no surprises.
Un email curat pe săptămână — profile noi de plante, note sezoniere de cultivare, adăugiri în arhiva de remedii. Gratuit, fără tracking, dezabonare oricând.