Privacy Policy
The new worksheet code uploads and persists no entered levels. Normal hosting requests and public search URLs have separate metadata boundaries; no blanket absence-of-logging or organizational privacy audit is claimed.
What we don't collect
- No accounts. No login.
- No analytics cookies.
- No tracking pixels.
- No third-party advertising tags.
- No affiliate-tracking cookies.
Server logs
The hosting provider (Cloudflare Pages) maintains standard request logs (IP, user-agent, URL, status, timestamp) for operational and security purposes only.
Email correspondence
TODO(Darren): confirm a monitored publisher correction/contact route. The previous draft inbox and response/retention promises are not authenticated here.
Children
FTM does not knowingly collect data from children under 13.
Public reading and private worksheet entries
The vintage bridge runs in the browser on declared levels. It asks for no account identity, credentials, transcript or private market dataset. Application code uploads and persists none of its entered values. Reset and reload restore the invented teaching case. The article claim-check form reads only a user-selected import and produces local exports. Do not paste confidential source material or personal identifiers into an educational workflow. A client-side worksheet is not a secure newsroom source system, verified anonymity service or legal guarantee of data handling.
Search and hosting have their own boundaries
Publication search uses visible text and may put the query in the URL, browser history and ordinary hosting request metadata. That is separate from the bridge inputs. Cloud infrastructure can receive normal requests even without an analytics pixel, so the page does not claim no logging occurs anywhere. External references lead to other operators with their own policies. This expansion adds no trading connection, advertising tracker or automatic API writer. A complete organizational privacy notice still needs the responsible publisher and actual provider arrangements rather than a blanket promise based only on static code.
An export is neither private nor authenticated by its format
JSON records preserve entered assumptions and limitations. CSV imports or downloads can contain user-supplied notes, so the retained workbench escapes formula-like spreadsheet cells. Malformed, incompatible or duplicate-identity imports should not overwrite an existing draft. A version field cannot establish the truth of the imported source. Store sensitive material using an appropriate authorized system outside this publication. The fictional dossiers contain no real interviewee identity or proprietary observations, and cannot be relabelled as anonymized actual reporting merely because their fields are complete.
Publisher records remain outstanding
The prior draft described monitored inboxes and correspondence retention without supporting operational records. This edition requests a confirmed contact rather than invent a working response process. TODO(Darren): confirm legal publisher, privacy responsibilities, applicable retention and monitored correction/security routes. Source rights and privacy minimization also require separate agreement in an actual investigation. Preserve the fields necessary for a scoped calculation without publish unnecessary personal data. A source gap should remain explicit, not be hidden by a confident policy statement or replaced with plausible confidential information.
Minimize the source before exporting it
A source record should contain only the fields necessary for the stated question and permitted by its owner. A working export button is not permission to redistribute a private transcript, respondent list or account identifier. Keep authentic confidential material in an appropriately authorized system and use only safe references here. The authored dossiers avoid real personal information; they remain fictional education rather than anonymized customer reporting.