Small inputs, clear boundaries.

The analyzer request can contain a public TikTok URL, audience, optional product context, goal, available filming material, transcript or on-screen text. If you request a revision, we also receive the selected draft reference, change reason and correction choices. Copy and review events record the analyzer surface, report and draft references, draft number, judgement and selected change reason. Revision records link the new report to its parent. These events do not send the draft text or private rewrite context and do not confirm adoption, filming or results. Anonymous product events contain a random visitor ID, a short session ID, the page path, coarse UTM values and event properties. The optional hook game records anonymous visits, choices, completion, copy actions and clicks to the analyzer. It does not send the nickname, share text or full page URL. When Google sign-in is available and you choose to use it, we receive your Google account identifier (subject), email address, name and profile picture URL, where supplied. We do not ask for passwords, payment details, private account cookies or contact lists.

Public-link metadata is fetched through a protected server-side connector. In manual text mode that lookup is skipped. When you request model-assisted suggestions, your supplied text and product context, available post description and published subtitles are sent to the configured model provider through our server. A requested revision uses the original report context and your explicit correction choices. The browser never receives provider keys. Do not paste secrets, private customer data or confidential source code into the form.

The optional AI tools on guide pages send your selected product category and audience, plus an anonymous visitor identifier for allowance protection. Text entered into those guide pages stays in the browser to fill the product-name slot. For a fresh three-draft analyzer result, an AI filming check starts automatically: the drafts’ opening lines and filming notes, and the associated product, audience and available-material context, are sent through our server to the model provider. This check uses additional free allowance and is separate from the original generation. If you choose to change a shot, your selected material and draft context are sent for that requested revision. These proposals do not verify product claims, material rights or your ability to film. The confirmed-product-page script flow is a separate process and does not automatically start this three-draft check.

The prototype stores browser visitor identifiers, server-side daily usage counts, hashed request identifiers and short event records for funnel measurement. Research logs record a hash of the query or source link and the operation status. Limited report diagnostics record service status, attempt counts, timing and available usage totals, without source text or your brief. Public video descriptions, counts and published subtitles may be cached and reused for up to 24 hours; that cache does not contain your audience, product context, supplied text or generated advice. Daily allowance protection uses a server-side keyed hash of the request IP and anonymous visitor ID; raw IP addresses are not written to these usage tables.

When revision is enabled for a completed report, we store an encrypted snapshot of its original input, source text, brief and drafts, with a 64 KiB limit and a 24-hour expiry. Short-lived encrypted revision results allow the same submission to be retrieved without starting another model request. Expired content cannot be used for revisions and is removed during subsequent cleanup requests. A secure, HTTP-only cookie binds anonymous revision access to the browser that received the report. Private rewrite references stay out of analytics, URLs, copied drafts and Markdown exports.

Operation references, hashes, limited correction choices, allowance records and execution status may remain after content expiry to prevent duplicate execution and reconcile uncertain results. They do not contain the source text, private brief or draft body.

When confirmed-task preparation is enabled, a submitted product-page link lets our server read available public page text and send relevant passages to the configured model provider to prepare a task for your review. Source, proposal and confirmed-task content is stored encrypted and is valid for no more than 24 hours from the original source capture. When free previews are enabled, confirming the task requests one model-generated preview of the confirmed product statements. That same request may include the accepted reference quotes and their expression labels to explain the proposed opening. The explanation is AI advice, not verification of audience fit or performance; no additional explanation request is started. Only a validated explanation is retained with the encrypted preview and saved task; unavailable explanations are not filled in later. Its encrypted result expires no later than that task; uncertain execution is kept as a status record to prevent another model request. Preview generation does not require Google sign-in or a purchase. Expiry prevents further use; clearing stored content depends on later cleanup, and user-linked operation references and status records may remain.

Customer-task measurement uses a random measurement reference to connect a confirmed task, its preview, saved-account readback and copy or download requests. These measurement references do not authorize access to your task. The short records contain finite action names, event and action identifiers, source classification, status, timing and available usage totals; they do not contain product text, generated scripts, image bytes, Google profile fields or private task-access references. Copy success describes the browser clipboard API response; a download request does not confirm a saved file or adoption. Measurement references and status records may remain after the original task expires; this does not extend access to its content.

When customer product-image capture is enabled, the server may retrieve one static JPEG, PNG or WebP image linked in the public product page, up to 256 KiB. Captured image bytes are stored encrypted, bound to the original task and accessible only through its authorized browser or saved account. They expire with the original task within 24 hours; opening or saving a task does not extend that expiry. Confirmation asks whether you can use the displayed image. The model receives the permitted material description rather than the image bytes; this does not perform image interpretation or OCR, verify ownership or establish product results. Expiry prevents further access and does not promise immediate physical deletion.

When task continuation is enabled and you choose to keep a task, browser session storage holds opaque task references, operation identifiers, consent, status and expiry. These references are valid for up to 30 minutes. It does not hold your Google profile, source text or brief. Canceling the continuation does not delete the server task.

When Google sign-in is available, your Google account identifier identifies a returning account. Your email, name and profile picture URL are used to display your account. Saved reports and tasks are associated with your account only when you choose to save them. These profile fields, account creation and update times, and a hashed app session token are stored in the account database. Google access tokens are used temporarily to retrieve your profile and are not persisted by this sign-in implementation.

The browser may load your profile image from its image host without sending a page referrer. Your Google account identifier, email, name and profile image URL are not automatically included in model prompts, public-link lookups or product analytics.

If you save a report, its source link, summary and report payload, including any supplied product context, filming material, transcript, screen text and generated drafts, are stored in account history. We do not currently apply an automatic deletion deadline to Google profile records or saved report history. The 24-hour public-evidence cache and task expiry do not apply to that history.

App sessions expire after 30 days. Signing out ends the current session; it does not delete the profile stored by AngleCue, saved reports or server tasks. Self-service account and report deletion is not available yet.

The buyer hesitation map receives the category, price band and keyword or product link you choose. Our server fetches public TikTok videos, comments and reviews for it through a third-party data provider and sends public comment text to the configured model provider to sort it; the browser never receives provider keys. A finished map is kept for 30 days so you can reopen it, linked to your account when you are signed in or otherwise to a random browser cookie.

The creator sample check receives the creator handle and the choices you make. Public profile and post statistics are fetched the same way and cached for up to 24 hours under a keyed hash of the handle; the cache and our usage counts do not store the handle itself. If you track a sample, the creator handle, product name, sent date and the result you report are stored with your account, and our server schedules checks of that creator's public posts during the 21-day tracking window. Checks may be delayed or unavailable; you can stop tracking earlier.

AngleCue Pro is sold through Paddle (paddle.com), the merchant of record. Card and billing details are entered on Paddle's checkout and are never sent to AngleCue. We receive and store your Paddle subscription and customer identifiers, subscription status and billing period dates, linked to your AngleCue account, and we count your Pro usage per billing period. Paddle receives your email address to create the checkout and send receipts; its own privacy policy applies to the payment.

Contact support@anglecue.app privately about your account, saved work or privacy questions. Please do not send passwords, access tokens or payment details.