Our commitments
These are not policies we try to follow. They are rules written into the code that runs the service, so a client's request, a payment, or a clever instruction hidden in a web page cannot change them.
What we never do
| We don't | Why |
|---|---|
| Automated replies, DMs or messages to strangers | Platform automation rules forbid unsolicited automated replies; they draw spam reports and suspensions. |
| Bulk cold email or cold texts | Email and texting law; texts need the recipient's prior consent. |
| Mass-generated pages to manipulate search rankings | Google's scaled content abuse policy, whoever or whatever made the pages. |
| Fake or AI-written reviews, bought followers or engagement, paid or suppressed reviews | The FTC rule on reviews and testimonials, with civil penalties per violation. |
| Guaranteed rankings or results | Nobody controls the platforms' algorithms; a guarantee is a false claim. |
| Storing a client's passwords or signing in as them | Official account connections only; passwords break on two-factor login and make us the breach target. |
| Scripting a platform's website to get around its API | Named as off-limits in platform automation rules; assumed everywhere. |
| Unreviewed content in finance, health or legal | Regulated claims carry legal exposure; an attorney sets the rules for each vertical first. |
| Hiding that an account, message or chat is automated where disclosure is required | Platform rules and consumer trust. |
| Photos of minors or identifiable customers without written consent | Privacy and safety. |
| Political or issue advertising | Extra verification and election rules; out of scope. |
| Code that rewrites itself in production | One bad change would hit every client at once; code changes stay human-reviewed. |
How AI is used, and disclosed
- Plans, drafts and reports are produced by software built on Claude. Your business profile is the only source of facts it may use; a draft that states a price, a phone number or a claim not in your profile is stopped before anyone sees it.
- Every item passes a check in code (the list above, false-claim patterns, your off-limits topics) and a read against the platform's current rules, fetched from the platform's own pages. If the check cannot run, the item waits for a person. It never defaults to publishing.
- Where disclosure of automation is customary or required, it is included and cannot be switched off. Our audit reports say when the model wrote them. Any chat on your site will say it is automated.
- The code does not rewrite itself. Plans, prompts and budgets adapt to results; code changes are reviewed by a person.
Your accounts and data
- We never hold your passwords. Connections are made on each platform's own login page and can be revoked by you at any time.
- Instructions are accepted only from the people you name. Anything else is filed and shown to a person, never acted on.
- Every action, its inputs and the reason are kept in a log that cannot be edited or deleted.
- Only your own photos and footage are used, and never identifiable people or minors without written consent on file.
- Leaving revokes every connection and hands you your content and data.
Money
- Flat fees for ad management; we never earn more by spending more of your money.
- Your ad spend never passes through our accounts.
- No guaranteed results, in writing.
Nothing on this page is legal advice. The responsible party for what the service does on your account is us, under the agreement.