Tools
Three interactive tools over the same catalogue the rest of the site reads. Each one states its limits on its own page rather than in a disclaimer at the bottom.
Tea finder
Filter 137 teas by family, flavour, aroma, mouthfeel, intensity, complexity, appreciation level, brewing style and experience level.
- You give
- Any combination of the nine facets. Nothing is required — with no selection the tool lists the whole catalogue.
- You get
- Matching teas with a 'recommended because' list drawn from the tea’s own authored fields, its flavour profile, its brewing styles, and links to the full record.
Only filters on fields every published tea carries. Caffeine is excluded because just 11 of 137 teas have a measured estimate and a filter on it would silently hide the rest; umami is excluded because it is authored mainly for shaded Japanese greens and would tilt the intensity model toward one origin. There is no health, sleep, energy or digestion filter and no field that could support one. Results are never scored — ordering is by how many of your selected values a tea shares, which is stated on each result.
Tea comparison
Put any two of the 137 teas side by side across category, production, origin, flavour, brewing, cultural context and who may prefer each.
- You give
- Two teas.
- You get
- A seven-row table with a cell for each tea in every row, the authored comparison where one exists, and the note explaining why the two are commonly confused where that is recorded.
There is no verdict, no score and no winner — the purpose is understanding a difference, not resolving it. The 'who may prefer it' row runs the same sentence structure over both teas so neither can be framed as the safe choice. Where the catalogue records nothing for a tea, the cell says so instead of inferring: eleven teas have no origin recorded because a blend has components rather than a place, and guessing a country from a name would be undetectably wrong.
Brew assistant
Temperature, leaf weight, ratio, full infusion schedule and equipment for a specific tea in a specific brewing style.
- You give
- A tea, a brewing style, and a vessel volume.
- You get
- Water temperature, leaf weight for your vessel, the ratio it came from, the infusion schedule, rinse guidance, equipment, and an explicit list of what the parameters do not tell you.
Never derives one brewing style’s parameters from another. Choosing a style with no authored parameters for that tea returns nothing at all, names the styles that do exist, and explains why nothing was calculated — scaling a western ratio into a gongfu answer produces a confident, plausible, wrong result the reader cannot detect. Eleven of fifteen described brewing methods, including kyusu, cold brew and herbal infusion, have no authored parameters on any tea and are listed as such. Family-level ranges report an observed spread and are never presented as a setting for any individual tea.
What none of these do
No tool asks what you want tea for. There is no field in the model recording what a tea does to a person, so sleep, energy and digestion cannot be filtered, recommended or implied — not as a matter of restraint but because the data genuinely does not exist. It arrives with an evidence layer that has its own human review, and that review is not signed.
No tool invents a number either. Where a parameter is missing, the tool says which one and stops — a scaled guess would be indistinguishable from an authored value on the page.