Privacy Policy

Information we process

We process account identity and contact details, Organization membership, registration and access status, newsletter subscription and delivery preferences, submitted product and launch content, Promotion Credit and market activity, administrative audit records, support communications, benefit fulfillment, and advertiser-reported result proofs. Stripe processes payment-card details; we retain payment references, purchase amounts, status, refunds, and disputes rather than full card numbers. Uploaded media and storage metadata are processed to display submitted content. Resend processes email delivery, and authentication uses cookies and session records.

Clicks and network information

Controlled outbound links record the destination entry, time, campaign parameters, verification state, referrer domain, coarse country and device classification where available, and privacy-preserving visitor identifiers. The structured click records do not store raw IP addresses or full User-Agent strings. Visitor hashes are scoped to a Market Entry and UTC day and are not intended as durable cross-site identities. Generic forwarded-IP headers are not treated as authoritative; deployments must explicitly select supported trusted provider headers or safe no-network-identity mode. Standard hosting, network, security, and diagnostic systems may separately process or temporarily record network and device information when operating and protecting the service.

Why we use information

We use information to provide and secure AttentionWatch, authenticate users, administer account access and explicit denials, process payments, maintain immutable financial records, operate and reconcile the Attention Market, detect suspicious activity, fulfill earned benefits, respond to support and legal requests, return requested discovery results, maintain service reliability, detect errors, enforce rate limits, prevent abuse, and investigate security or operational incidents. Plausible may process privacy-oriented site analytics when configured.

Public API and compatible AI clients

When you or a compatible AI client use the AttentionWatch public API or MCP service, we may receive the requested tool or endpoint, task-specific input needed to perform that request, protocol and coarse client metadata, and standard network information needed to deliver and protect the service. Calling a tool does not, by itself, give AttentionWatch your complete ChatGPT conversation. A client may send task-specific content as an input; that client has its own privacy practices.

Our structured Public API telemetry records a request ID, endpoint, API version, coarse client family, duration, HTTP status, error class, result count, cache status, rate-limit outcome, and timestamp. Structured MCP telemetry records a request ID, tool name, MCP version, coarse client family, duration, success or error class, result count, rate-limit outcome, cache profile, and timestamp. These records help us operate, secure, troubleshoot, and rate-limit the service. They are not designed to record full ChatGPT prompts, complete chat histories, raw tool arguments, request or response bodies, authorization tokens, cookies, raw IP addresses, or full User-Agent strings. Application error, framework, hosting, and security logs may contain additional diagnostic information. OpenAI and other compatible clients process information under their own policies; see the applicable provider’s privacy documentation for its practices.

Sharing and processors

We share only what is reasonably necessary with service providers such as hosting, database, email, storage, caching, analytics, identity, and payment providers. Public submissions, rankings, verified click totals, historical results, and clearly labeled advertiser-reported proofs may be displayed publicly. We may disclose information when required by law or to protect users, the service, or legal rights.

Retention

Raw ClickEvent rows are normally retained for 90 days and then deleted in bounded batches after daily aggregate counts are preserved. Financial ledgers, allocation history, Stripe event records, administrative audits, reconciliation evidence, newsletter consent or opt-out history, account records, and other records needed for accounting, fraud prevention, disputes, security, or legal compliance may be retained longer.

Operational records are retained according to the system that produces them; there is no universal fixed-day deletion period for all logs. Structured API and MCP telemetry is emitted into application container logs. Application, PostgreSQL, and Redis container output uses size-based rotation, currently up to three 10 MB files per container, rather than a fixed age limit. Container replacement or removal may shorten live retention. Traefik also uses Docker container logging, but does not have that same explicit per-container size cap. Host system logs follow separate age, size, or weekly rotation settings. Some Dokploy-generated diagnostic and deployment log files currently have no established automatic age-based deletion. Historical copies may remain in rolling whole-server backups after live logs rotate or disappear, until those backups are replaced.

Your choices and requests

You may request access, correction, deletion, or export of applicable personal information, subject to identity verification and legal, security, financial-record, and backup exceptions. Newsletter subscribers may use the unsubscribe link in an email; signed-in users can also manage email preferences in their account. Contact cife@attentionwatch.com. We do not knowingly offer AttentionWatch to children under 13. International processing may occur where our providers operate. Depending on where you live, applicable law may provide additional privacy rights or requirements. We will honor applicable legal obligations and valid privacy requests as required by law. Material changes to this Policy will be posted here with a revised effective date.

Effective: September 24, 2026.