Enter two atproto handles. This reads both accounts' real public profiles and last 100 posts — nothing private, nothing inferred about who anyone's actually into — and turns the patterns into a hypothetical, pop-psych-flavored compatibility reading: timezone sync, chaos compatibility, communication style, humor alignment, shared vocabulary, interest overlap, acts of service, whether they're already following each other, plus a love-language and attachment-style read for each account. It's a toy, not a diagnosis — see "how does this work" below.
Each handle's public profile and last 100 posts-and-replies
(app.bsky.actor.getProfile +
app.bsky.feed.getAuthorFeed), plus the real follow
relationship between them (app.bsky.graph.getRelationships),
feed eight scored dimensions:
The headline score is a weighted average of all eight. The "love language" read scores five real ratios (exclamation rate, alt-text rate, reply rate, media rate, emoji rate) against a generous reference ceiling and reports the highest as each account's primary language — five numbers per account, not one coin flip. "Attachment style" is reply rate and posting-rhythm irregularity mapped onto four familiar-sounding labels, purely for flavor.
Nothing here reads private data, DMs, or likes, and nothing here claims to know who anyone actually has feelings for — it's a hypothetical "what if" read off public posting patterns, framed as pop psychology because that's more fun than a spreadsheet. Same two handles, same public data, always the same reading, until one of them posts or follows again.