Frame Mirror
A private, structural record of the texts you read. Which frames they use, which perspectives they leave out, how your reading mix shifts. Browser-generated ID, no content stored, no email collected.
What Frame Mirror is
Frame Mirror keeps a structural log of the texts analyzed from this browser: which frames were detected, which perspectives were missing, the voice classification. After a handful of analyses, the log surfaces patterns in what you read and where analytical dimensions consistently drop out. The underlying measurement is the same one Frame Check uses; the Mirror is the slice of it that stays on this browser.
What is stored: a browser-generated ID (not your email), plus counts and categorical summaries of each analysis (which frames were detected, which perspectives were missing, the voice classification). If you also use the optional decision-tracking and frame-disagreement affordances, the free-form prose you type into those forms (decisions, outcomes, disagreement reasons) is stored verbatim, capped at 500 characters per field. The privacy page enumerates the schema in full and explains how we read aggregate signal from these tables. What is not stored: your texts, your queries, any text you pasted, any personal identifier beyond the anonymous browser ID. Export or delete everything at any time.