Data-Use Disclosure

Trace sends the prompt, user request, selected source excerpts, and task metadata needed to complete an AI task. Examples include strategy generation, source extraction, voice scoring, signal assessment, and draft generation.

Trace does not send an entire connected GitHub account by default. It works from selected repositories, normalized activity, and source chunks relevant to the task.

GitHub access is used to connect your account, identify authorized repositories, sync selected repository activity, and keep source citations attached to generated story seeds and drafts.

The launch goal is explicit user-controlled repository selection. Trace should never scan every repository by default when a narrower selected-source flow is available.

When you connect Reddit, Trace requests standard identity, read, and submit OAuth scopes. Trace scans public discussion threads in communities you monitor to discover conversations relevant to your verified product facts.

Trace does not store full user histories or private messages. Excerpts are processed by AI models solely to rank relevance and draft helpful responses on demand. No automated bulk posting or DM automation is performed. Every reply or standalone post requires explicit user review and individual approval before publication.

You can pause scanning, pause publishing, or completely disconnect your Reddit account at any time in Settings.

Stripe handles payment methods, checkout, invoices, subscription status, billing portal access, and webhook events. Trace stores Stripe customer and subscription identifiers plus reconciled subscription state so feature access can match the billing source of truth.

Trace breaks uploaded or connected source material into chunks so the product can retrieve specific evidence and cite where a draft came from. Chunks may include text from files, commits, pull requests, README content, or user-provided answers.

Chunks are tied to your account, provider, and source identifiers. They exist to make content generation specific and auditable, not to train foundation models.

Trace measures activation and reliability events such as signup, strategy generation, source connection, sync completion, story seed review, content approval, scheduled content, queue failures, webhook delays, and AI budget exhaustion.

These events help the founder support launch users, detect outages, and confirm Phase 2 success metrics are measurable from product behavior.

A full account export includes your profile, connected-source metadata, source chunks, story seeds, generated content, calendar, strategy doc, check-ins, and a monthly summary of AI usage and cost — never your password, OAuth or source access tokens, or per-call AI prompts.

Uploaded files are included as time-limited (1 hour) download links generated at export time, not permanent public URLs.