Blog Smith — Protected Customer Data

Blog Smith is Protected Customer Data (PCD) Level 0. It does not request the read_orders or read_customers scopes, and it never reads, stores, or processes any order or customer personal data.

What Blog Smith accesses

• read_products — product and collection handles, used as internal-link slots and grounding facts for content outlines.
• read_content / write_content — your store blog, to publish approved, scrubbed articles you have reviewed.

What Blog Smith stores

• Content outlines (topic, target keyword, section word targets) you create.
• Article drafts produced against those outlines, until you publish or delete them.
• Aggregate token-usage counts for billing — never prompt or response content.

No orders, no customers, no personally identifiable information of your shoppers is ever touched.

GDPR / compliance webhooks

Blog Smith implements the mandatory customers/data_request, customers/redact and shop/redact webhooks. Because no customer data is held, the customer webhooks truthfully report that there is nothing to return or redact; shop/redact deletes the store’s Blog Smith data — outlines, drafts, usage counts, web-vitals samples and the access token — except one record kept on purpose: a salted, irreversible hash of the shop domain plus the date its one free AI-writing allowance was issued, held as an anti-abuse measure so the allowance cannot be re-claimed by reinstalling. It contains no shop name, no contact details and no customer data, and cannot be reversed into the domain.

Honesty

Article length and keyword targets are predicted heuristics, sized to the competitor word counts you enter — not a guaranteed ranking. AI-written drafts are strictly grounded in your own catalog and pass a de-AI scrub (watermark stripping + AI-tell flagging) with a hard block on overclaims, and are published only after you explicitly approve them.