What this policy covers
This policy describes the current Midnight Windows application, its connection service and support contact. It explains the information used to organize your work and carry out the features you choose. Connected platforms, AI providers and hosting providers also process information under their own policies and settings.
Information on your computer
Midnight stores creator settings, imported file references, titles, captions, hashtags, schedule records, account identifiers, cached platform information and activity history. Original video files remain where you keep them. Importing a video does not, by itself, upload it.
Workspaces organize separate creators’ records within the app. They are not separate Windows security accounts. Designated credential stores and protected records use Windows Data Protection API (DPAPI) protection for the current Windows user. This does not encrypt every source video, setting, export or activity log. Protect your Windows account and backups.
Account connections and the connection service
Supported sign-in flows use the platform’s authorization screen. Midnight does not need to collect your social-media password. The app receives authorized tokens and account information for the features you allow. Local account credentials are kept in designated protected storage.
The hosted connection service handles authorization codes, proof-bound sign-in handoffs and token exchange or refresh requests where a platform requires a confidential application secret. Tokens pass through that service for these operations; its current handoff state is short-lived and held in memory rather than a durable customer-token database. Platform application secrets remain on the service, not in the desktop distribution.
The connection service currently uses Render free hosting. Idle sleep, restart or deployment can interrupt an unfinished handoff. Hosting infrastructure may process IP addresses and request metadata under its own configuration and policies. Application code avoids intentionally placing credentials or video contents in logs; this is not a promise that every hosting-provider record has the same retention.
Optional AI assistance
If you enable and use an AI provider, Midnight can send the selected request, relevant creator/video/calendar metadata, limited conversation context and sampled JPEG frames from selected videos to that provider. A visual review examines sampled moments; it does not establish continuous-video or audio analysis. If automatic or batch reviews are enabled, configured reviews can send frames and metadata for selected imported videos without a separate prompt for every request. You can turn that feature off in the app. Review the information you enter, because captions and chat messages can themselves contain personal information.
Provider requests use your configured credentials and provider settings. Provider billing and usage charges may apply. Midnight does not include permission to spend money beyond the features and account settings you authorize. Provider retention, processing locations and account controls are governed by that provider’s policies.
Publishing and verification
When you authorize an available upload or publishing workflow, Midnight sends the selected media and reviewed metadata to the intended platform. An approved app-managed job may do this at its configured execution time. Platform processing, moderation, visibility and retention are then controlled by that platform.
Supported reads retrieve account identities, media, metrics or schedules where the platform exposes them. Public TikTok media is different from its native future queue; Instagram app-managed delivery is different from a native Business Suite schedule. Midnight does not treat an unverified local plan as a confirmed platform result.
Updates, support and retention
Update checks download public release information and application packages over HTTPS. An update check does not require a social-media login or upload your video library. The hosting provider may receive ordinary download request metadata.
If you contact support, your message may include contact details and diagnostics you choose to provide. Send only what is needed and remove passwords, access tokens, API keys and private account details from screenshots. Support information is used to respond and maintain relevant support records.
Voluntary in-app issue reports
If you choose Send report in a build with a connected Issue Desk, JR Studio receives the text you enter and any diagnostic details you choose to include. The preview can show the app screen, action, build, platform and an error code. A previous crash report can include its timestamp and error code. You can turn off the diagnostic details and send only your message. Review free text yourself and remove passwords, access tokens, API keys and private account information before sending.
JR Studio uses reports to respond to problems and test fixes. If an authorized operator starts the private fix workflow for your ticket, its free-text message, included diagnostics and staff triage notes can be sent to GitHub Actions and processed by OpenAI Codex to investigate, propose and test a change. The Issue Desk and these services may each process support records under their applicable settings and policies. You can contact JR Studio about information in your report using the address below.
Local records can remain until removed or the applicable app data is deleted; backups can retain earlier copies. Removing a local record does not delete the original file, a platform post or information held by an AI/hosting provider. Disconnecting locally and revoking authorization at a platform are separate actions.
Your choices and contact
You can choose which files and accounts to connect, whether to use AI, and which available operations to authorize. You can stop app-managed work, disconnect accounts and revoke platform access through the platform’s controls. Contact jrstudioapp.support@gmail.com for help locating local records or requesting assistance with information held by JR Studio. Do not include secrets in the request.
Material changes to this policy will be reflected in its published policy date. Review the current version when using a newly introduced feature.