Privacy, in plain language.

1. Data collected

The free OpenAI plugin receives exactly three creative briefs containing your app name, benefit, hook, body, call to action, template, accent color, and optional disclosure. Its prepare_studio_batch tool returns those briefs inside a shareable Studio URL, along with first-party sample links, helper links, and review requirements. The server processes the briefs to make the link; it does not save the briefs or link in ReelParcel’s application database. This tool requires no account and receives no source media.

In the free Studio, screenshots, recordings (including any source audio), edited copy, media selections, and review confirmations are processed in your browser to make downloadable creative files. Source media is not uploaded by this browser export flow. Explicitly saved drafts and named projects keep media and edits in that browser’s device storage. The optional local helper also handles local file paths, project names, export settings, downloaded files, checksums, and export-status receipts on your computer.

If you sign in or use authenticated workspace tools, we handle your account identifier, email and sign-in profile, session credentials, saved product facts, audience, tone, destination URL, and creative defaults. If you choose GitLab sign-in, we receive profile details, including an email address when available. Where enabled, reviewer sign-in sends the entered email and password to Supabase for authentication. Agent connections use permissions, token hashes or OAuth grant records, expiry and revocation details, and usage limits.

When hosted features are available, we handle uploaded screenshots or recordings, filenames, labels, dimensions, size, type and duration; creative batches, copy, captions, alt text, revisions, fingerprints and approvals; and quotes, credit reservations, render progress, errors, output metadata and resulting files. Authenticated MCP tools can return saved facts, creative specifications, asset metadata, temporary private media links, and output manifests to your connected client. The separate preview_asset tool can send a resized screenshot inline to that client. The directory connector’s saved-output tools receive a render identifier and return status, output links, or a manifest containing approved copy and specifications.

Optional browser analytics collect page views and bounded events about imports, previews, saves, exports, downloads and failures, with template, format, draft-source and coarse referral or entry categories. Events may include a random browser-session identifier, a random output or event identifier, optional feedback verdict and reason, and your account identifier when signed in. The first-party product-event fields exclude media, creative copy, filenames and raw referring URLs. Successful free remote tool calls may separately record a timestamp, event name, classification and fresh random identifier without briefs, links, media, account ID, IP address or user agent in that event record.

Hosting and authentication services also process network and request information, such as IP addresses, requested URLs, browser information and diagnostic errors. Error logs may contain provider error details; hosted worker logs include job identifiers, timing, resource use and failure information. If paid checkout is enabled, Stripe collects payment details and ReelParcel records customer, checkout, payment and refund identifiers, amounts, currency and credit history. If you email support, we receive your address and whatever you include in your message.

2. Purposes

We use briefs to prepare editable Studio handoffs, device media and edits to create files you request, and saved workspace data to let you review, revise, reuse and retrieve your work. Hosted rendering uses your selected media and approved creative specification to produce exports when that feature is available. ReelParcel’s composition and handoff tools use deterministic templates; they do not call an AI inference service to generate or analyze your content.

We use account and connection records to authenticate you, enforce ownership and permissions, honor revocation, limit requests, and prevent duplicate operations. Payment and job records support credit accounting, fulfillment, refunds and recovery from failures. Analytics and diagnostics help us understand workflow use, find errors and improve the service; support messages let us respond to your request.

3. Recipients and processors

Vercel hosts the canonical website and remote MCP endpoints and receives requests and server diagnostics; Vercel Web Analytics receives enabled page-view and product-event data. Supabase provides account authentication, the application database, private hosted storage and signed media links, and stores first-party product events. The Fly.io worker reads hosted job specifications and source media, renders files, uploads results to Supabase, and reports job progress to ReelParcel. Netlify is retained for the legacy hosting address and rollback deployment; requests to that address are handled there before any redirect.

GitLab handles GitLab sign-in when you choose it. Stripe handles checkout, billing management and refunds when paid features are enabled; ReelParcel sends account/email and checkout metadata and receives payment status, rather than collecting full card details in its own forms. Our Gmail support inbox uses Google to receive and retain messages you send.

When you use ChatGPT, OpenAI receives the tool arguments and results in that conversation, including the copy-bearing Studio link. Other MCP clients you choose likewise receive their tool inputs and results, including private links or screenshot pixels when the selected tool returns them. Their processing, retention and any model-training choices follow that provider’s terms and your settings; ReelParcel cannot erase their copies. ReelParcel does not send these workflows’ briefs or media to an additional AI inference provider or to PostHog.

Anyone you share a free Studio handoff URL with can read its embedded copy. Keep private download URLs private: a recipient can use a signed URL until it expires. Files you download or share, and local paths or receipts returned to a local MCP client, also become accessible to the recipients you choose.

4. Retention

The free remote handoff has no saved application record of your briefs or Studio link. The link itself carries the copy and can remain in your chat, clipboard or browser history; Studio removes the fragment from its current address after import. Saved device drafts, named projects, helper browser profiles and downloaded files remain until you delete them or your browser or operating system clears them. Local export-status receipts become unavailable ten minutes after an export completes or fails while the helper is still running. If a review is interrupted, its local status receipt remains available until the original review deadline plus ten minutes; receipt files are cleaned on later helper access or activity, rather than by a guaranteed background erasure.

Saved account, product, revision, approval, job, payment and connection records have no general automatic deletion deadline in the application. They remain until removed through the applicable project or account deletion process. Expiring a credential, quote or download link ends its use; it does not by itself erase the underlying record or file.

Hosted media retention

Hosted media may have a configured expiry shown in the workspace; expired media stops receiving new preview or download links. Existing media without a registered expiry has no automatic expiry applied retroactively. Physical removal depends on cleanup running and outstanding upload or render holds clearing, so a download expiry is not a promise that all stored bytes have already been erased. We do not promise a fixed automatic hosted-file deletion period.

First-party product events have no automatic expiry and may remain indefinitely. Account deletion removes the account link from those events rather than deleting the events themselves. Free remote success counts have no saved account link to remove. Provider logs, backups, analytics, payment records, support messages and copies held by your AI client follow the relevant provider’s retention and deletion processes; we do not promise a single fixed deletion period for those copies.

5. User controls

Review and edit your copy before export, and provide only media you own or have permission to use. Keep secrets and private details out of briefs and handoff links. Download your browser exports in Studio, or retrieve available completed hosted outputs in the workspace or through authorized output tools. To request a copy of account data beyond those downloads, email nestlingapp1@gmail.com; there is no self-service full-account archive export.

Use the data deletion page in each browser to clear its saved Studio drafts and named projects. “Forget saved media + draft” clears only the current draft. For the local helper, run node reelparcel-free-export.mjs --profile-location, close its browser, and delete the reported profile directory; delete downloaded outputs and the separate local-receipts directory beside that profile if you also want those local copies removed.

When signed in, the data deletion page offers account deletion with an explicit DELETE confirmation. The service removes your private-storage prefix, deletes the associated Stripe customer when configured, and deletes your authentication account and cascading workspace records. Active renders, unexpired upload links or unresolved checkout payments can block deletion until resolved. Project deletion can leave media cleanup pending, so deleting a project is not immediate proof of physical media erasure.

Browser analytics start enabled unless you turn them off in the footer or send Global Privacy Control or Do Not Track signals. A separate remote MCP client must send DNT: 1, Sec-GPC: 1, or X-ReelParcel-Analytics: off to suppress free-tool success counting. The browser preference does not reach that client, and opting out does not erase earlier events or essential hosting logs.

Use Integrations to revoke personal API tokens or OAuth agent connections, and manage conversation copies and privacy settings with your AI provider. Send privacy, access or deletion questions to nestlingapp1@gmail.com, including enough information to identify the relevant account or issue. Do not send passwords, access tokens or private source media by email.