Privacy and account control

Updated November 6, 2026

This notice describes the current owner-operated Fabrik experiment and its YouTube integration. The operator is Will Clark, reachable at will.clark.codex@gmail.com. It does not describe a public, multi-user service or a book seller's payment terms.

Data the connected app uses

• Authorization: OAuth access and refresh tokens, the permissions granted, and the selected channel's ID and display name.
• Content and publication: owner-supplied video files and metadata, exact content hashes, reviews, publication IDs, URLs and timestamps.
• Available results: video counters and authorized Analytics reports, such as views, watch time, engagement, subscriber changes and reported search referrals. A field the provider does not supply remains unavailable.
• Workspace records: the owner's notes, pins, research sources, decisions and operational history.

The current integration does not request Gmail, Drive, contacts, monetary Analytics, or viewers' personal profiles. It does not infer a named viewer's identity from aggregate reports.

Why the data is used

Fabrik uses these records to verify the correct account, carry out authorized publication, prevent duplicate posts, show measured results, recover interrupted work, and inform the owner's next content decision. It does not sell Google user data or use it for advertising targeting. The integration has no model-training feature.

Storage and service providers

Cloud credentials are stored separately from report exports as encrypted values in AWS Systems Manager Parameter Store. The worker uses private AWS storage and a cloud database for operational records, media and measurements. The operator may retain local working copies and reports. Access is restricted to the operator and configured service roles.

Google receives authorization and API requests. YouTube receives videos and metadata when the owner authorizes publication; the chosen visibility determines who can see a post. AWS hosts the worker, private media and operational records. Selected footage, transcripts, publishing context and measured outcomes may be sent to the configured model provider, currently OpenAI, to generate or review content and inform editorial decisions. These data are not used by Fabrik to train a general-purpose model. GitHub Pages hosts these public app-information pages. The pages contain no advertising pixels, analytics scripts, embedded video or data-entry forms; the hosting service still receives ordinary connection requests.

Fabrik uses YouTube API Services. Google's handling of data is described in the Google Privacy Policy.

Revoke access or request deletion

You can revoke Fabrik YouTube's access in your Google account permissions. Revocation prevents future authorized API access; it does not itself erase copies already stored by Fabrik or remove videos already published to YouTube.

To request disconnection, a copy of retained data, or deletion of stored account data, contact will.clark.codex@gmail.com and identify the channel and request. Do not send passwords or access tokens. Published videos can be managed directly in YouTube Studio.

Current limitation: revocation does not yet trigger an automatic purge of journals, cloud archives or exported copies. Retention and deletion require operator action. This experimental setup does not claim a qualified automatic deletion lifecycle or authorization for indefinite storage.

Changes

This notice must be updated before the app adds new account data, recipients or uses. The current permissions and this notice do not authorize a broader integration.