Privacy

Scope

Detrás del Botón is developing a channel of animated explanations for adults about everyday systems. Its internal research and production tool runs on a Windows computer and is not offered as a public analytics service. The owner, acting as an individual, is the sole user of this internal client. This policy is presented to them locally.

The implemented workflow uses YouTube API Services to retrieve public video and channel metadata, including third-party metadata. Requests, refreshes and reuse of that data for judgments are paused. The pause does not imply deletion of earlier data or constitute authorization or certification by YouTube.

Data and purpose

The workflow retrieves titles, descriptions, video and channel identifiers, publication dates, language, duration and public statistics. It records search queries, acquisition dates, processing results and execution evidence. The purpose is to study existing coverage of topics and explanatory questions for the channel's editorial research.

The code includes comparative indicators, internal scores and its own relevance and coverage judgments. Their use with API data is paused and its admissibility remains under review. These outputs are not official YouTube metrics, measurements of audience demand or predictions of revenue or success. Describing these functions does not claim that they are authorized.

A manual record exists for the channel's observations after publication. An operational YouTube Analytics integration for choosing topics has not been demonstrated. This text does not claim that OAuth permissions have been granted or that private data is accessed under an authorization.

Destinations and processing by OpenAI

The judgment workflow uses OpenAI through Codex with ChatGPT sign-in. The implemented function sends selected title and description content with the instructions and project material needed for the judgment. The current pause blocks new judgments on that corpus through the protected entry points; it does not establish that content has never been sent.

Data and records are stored on the computer, in local Git and through GitHub backups. The current remote repository was verified as private. The owner stated that there are no additional copies outside this computer and GitHub. That statement does not replace verification of every copy, access, collaborator or historical visibility.

The owner stated that “Improve the model for everyone” is enabled and prefers to keep it enabled. The setting has not been observed directly in this assignment. OpenAI's documentation establishes that this control also applies to Codex tasks on personal plans; content may be used to improve models. This does not prove that specific content has been used. Using Codex with ChatGPT and OpenAI data controls.

OpenAI describes a separate control for training on full Codex environments. Its state and applicability to this workflow have not been verified; neither is inferred from the preceding setting. How data is used to improve models.

The retention periods and removal mechanisms that apply to executions already sent to OpenAI require clarification. No zero-day retention is promised, nor removal of earlier content by deleting a local copy or changing a setting. API or Business guarantees are not assumed solely because Codex is used with ChatGPT.

Retention, review and removal

The checked readers and adapters preserve the evidenced acquisition date and reject inputs whose date is invalid, missing, in the future or expired. The current preventive control uses a 30-day limit for evidenced API-data freshness; reading a cache or recovering an execution does not restart that date. This control does not establish coverage of all files or that their copies have been refreshed or deleted.

Physical retention remains unresolved. The owner has decided to keep a cache of 801 entries for now. That retention is not presented as admissible merely because reading it is blocked. Actions remain outstanding by copy, field and destination, including Git history, backups and the provider.

Refreshing or removing data requires locating the affected copies, determining the applicable action and checking its result. Deleting a local file does not by itself remove Git versions, backups or data retained by providers. Nor does it modify data held by YouTube. Complete removal is not claimed to have occurred.

Contact and requests

The owner is the client’s sole user and records decisions and requests for review or removal locally. To record a request, they identify the affected data or dataset and the requested action in the operation record. Before acting, the scope of copies, applicable obligations and verifiable outcome are determined; recording the request does not establish that removal has occurred.

The intended brand contact is contacto@detrasdelboton.com. Receipt confirmed on 20 September 2026. Handling of requests and replies from the brand address remain unverified. The owner does not need to email themselves to accept these documents or record their own decision. Passwords, API keys and Google credentials must not be included.

Google services and changes in scope

The client uses YouTube API Services. The Google Privacy Policy and YouTube Terms of Service are available for reference. The internal client terms are presented separately.

If authorized access to private data is added, this policy must describe actual permissions, revocation and effective removal before that processing is activated. The Google permissions page allows authorizations to be reviewed; this link does not establish that the client has one.

Material changes will require a new reviewed version and the applicable acceptance before the affected functions are enabled. The owner’s acceptance of P2/T2 has been recorded. It does not by itself enable access or lift the pause.

Local access to the policy

This policy and the linked terms are kept with the client’s records and must remain accessible to the owner before and during use. Local reading and acceptance do not require publishing a website. The ES/EN versions are a project deliverable, not a bilingual publication requirement attributed to YouTube.

These pages do not run scripts, forms or automatically loaded remote content. Official links open if the reader activates them. This document does not describe or enable public website services. Extending access to other people will require reviewing the practices and mechanism before granting access.