Privacy Policy
What information Pure City Research collects when you use purecityresearch.org or the Pure City Research plugin, and how we use it.
Information we collect
Cookies
Public analytics is currently configured without analytics cookies or local storage. WordPress may set essential cookies for signed-in administration, and—if comments are enabled—a visitor can opt in to a convenience cookie that remembers comment details. Other services may process data under their own policies.
Third-party services
The site is hosted on Cloudways and protected by Cloudflare. Koko Analytics and Cloudflare Web Analytics measure visits, FluentSMTP handles operational site email delivery, and search engines such as Google and Bing may index public pages. These services can process technical data such as IP address, browser information, page path, referrer, campaign parameters, and timestamps under their own policies.
Cloudways hosting-log retention. On 30 September 2026, Cloudways Support confirmed that Cloudways-managed Apache, Nginx and PHP access, error and slow logs are retained for up to 30 days. This includes rotated and compressed log archives. Logs older than 30 days are automatically deleted, and Cloudways does not keep managed log copies beyond this period. The dashboard's latest-1,000-record display limit is separate from this retention period. This period applies to Cloudways-managed hosting logs, not website backups, reader requests, newsletter records or other providers' records.
Reader-request storage. Request submissions are stored as private WordPress records and a notification is emailed to the site administrator. No automatic retention period is currently promised. You can ask us to delete a request by using the Contact page.
We do not sell personal data.
Your rights
You can ask for access to, correction of, or deletion of personal information associated with a comment, email subscription, or reader request. To make a request, use the Contact page. We may need enough information to identify the relevant record. Depending on where you live, applicable privacy law may provide additional rights and complaint routes.
Changes
We may update this policy. Material changes will be noted on this page with the updated date.
Last updated: 30 September 2026
Daily research email digest
If you subscribe to the Pure City Research daily digest, we use your email address, consent version and subscription/confirmation status to provide research updates. You join only after confirming through your inbox. We send at most one digest per day when new articles are available, and may send a confirmation message when you request to subscribe. Missed days may be combined into the next digest.
This list is separate from reader requests, Pure City Shop and SameCase registrations; those records are not automatically subscribed. The signup form does not ask for health information or record article-topic preferences. Please do not submit medical or other sensitive information through newsletter forms.
The newsletter stores subscription records and operational issue/delivery records. Short-lived signup-abuse counters use keyed pseudonyms derived from email and the network address; our hosting and email providers may separately process technical logs. The digest uses a dedicated connection to our configured email delivery provider. We do not add tracking pixels or track newsletter link clicks.
You can unsubscribe through the link in each digest. We retain the information needed to provide the service, record consent, honour unsubscribe requests and manage delivery; retention is reviewed periodically. Contact [email protected] for subscription or privacy questions.
Pure City Research plugin
The Pure City Research plugin is published by Pure City Life and retrieves public research from purecityresearch.org in ChatGPT or Codex. This section applies to that plugin in addition to the website practices above.
Information processed. The host sends selected research-tool arguments to the plugin service: a short ingredient or general research topic for search, or a numeric article identifier and optional text offsets for article retrieval. The service requests public content from this website and returns article text, cited source links, and available evidence, safety and review information to the conversation.
Authentication and hosting. The plugin is hosted through OpenAI Sites. Sites authenticates the connection and supplies a site-scoped user identifier, which the service checks for access. The current application code does not intentionally record that identifier or forward it to the research website. It does not create user profiles, save research queries, collect email addresses, add application analytics or provide a purchase flow.
Providers and retention. The research website can process requested public API paths and the search topic in its normal Cloudways hosting logs, subject to the up-to-30-day log retention described above. ChatGPT, Codex and OpenAI Sites can separately process conversation, request, authentication and technical information under their own policies, including OpenAI's privacy policy. The Cloudways period does not apply to those providers' records. The plugin application has no research-query database or saved health profile to delete; this does not mean that host conversations or provider logs are never retained. Manage host conversations and provider privacy requests through the relevant provider. A fixed retention period for OpenAI Sites technical logs has not yet been confirmed.
Use general topics only. Do not send diagnoses, medication lists, test results or other private medical information through the plugin's search tool. For plugin privacy questions or requests concerning information held by Pure City Life, contact [email protected] or use our Contact page. We may need enough information to identify a relevant record.
Plugin coverage and Cloudways log retention updated: 30 September 2026
Pure City Research Public — independent edition. This section applies to the separate public edition published by PureCity DWC-LLC, in addition to the relevant website provisions. It does not replace the preceding section describing the original OpenAI Sites edition.
Information processed. ChatGPT or Codex sends selected research-tool arguments to the public service: a short ingredient or general topic for search, or a numeric article identifier and optional text offset/length for retrieval. The service returns public article/methods text, source links and available evidence, safety and review information. It does not request health profiles or account information. Use general topics; do not send diagnoses, medication lists, test results or other private medical information.
Connection and provider. The approved public-test deployment is hosted through Cloudflare Workers Free. Production availability, resource limits and cache behavior have not been verified. Its three tools are anonymous; they do not require account linking or a Sites-authenticated identifier. The application does not inspect or forward incoming authentication credentials, browser cookies or host account identifiers to the research website. The reviewed application has no research-query database, account profiles, email collection, purchases or application analytics, and no deliberate tool-argument/request-header logging. Search topics are sent to the publication's public API and may appear in website/provider technical records. During the approved Workers Free public test, existing Cloudflare invocation logging and dashboard persistence remain enabled with 100% log sampling. The reviewed application does not deliberately log tool arguments or request headers. The exact content and retention of provider technical records have not been established for this test.
Retrieval limit. Each upstream response larger than 256,000 bytes is rejected explicitly; the service directs the reader to the public website rather than claiming an evidence conclusion from incomplete retrieval. The service can be unavailable when provider quotas or resource limits are reached.
Public-content cache. The application may cache a derived copy of a public article for up to five minutes, keyed by article number and software version, to reduce repeated retrieval and parsing. The copy contains public article text, source links and available review/safety/funding data. It does not contain search topics, user identifiers, incoming headers or conversation data. Retrieval timestamps identify the copy; provider caches may evict it sooner. This period concerns public article copies, not technical-log retention.
Providers and retention. Search sends the general topic to the website's public WordPress API; article retrieval sends an article number. The research website can process those API paths and search topics in its normal hosting logs. The existing Cloudways-managed log retention statement applies to those website logs, not Cloudflare Workers, OpenAI or other providers. Cloudflare can separately process technical/request data for delivery, operation, security and billing under its policies. ChatGPT and Codex process conversation and tool data under their provider policies. No fixed retention period for Cloudflare infrastructure/technical records has been confirmed for this deployment. Disabling application logging does not establish that providers retain no records.
Requests and deletion. This application's proposed implementation has no research-query database or saved user health profile to delete. That does not establish deletion of host conversations or provider logs. Use the host/provider's applicable controls for those records. For Pure City Research support and privacy questions, use the existing Contact page. Existing website reader-request, email and newsletter practices continue under their respective policy sections.