Acceptable Use Policy

You agree not to:

• Use the service for any unlawful purpose or in violation of any applicable law
• Upload content that is false, misleading, fraudulent, defamatory, obscene, or infringing
• Reverse-engineer, scrape, or attempt unauthorised access
• Resell or sub-license access to the platform without our written consent
• Use the service to harm minors, transmit malware, or send unsolicited communications
• Bypass rate limits, security controls, or audit mechanisms

Violations may result in suspension, termination, or legal action without notice.

Thmenu exposes a machine-readable Agent-Ready Menu Protocol (ARMP) via an MCP-compatible API endpoint. If you or your systems use ARMP or any other automated interface to interact with Thmenu on behalf of diners:

(a) Same rules apply. All Acceptable Use obligations in §4 apply equally to automated agents, regardless of whether a human is directly supervising the interaction.
(b) Rate limits are per identity. Each OAuth client identity (agent_subject) is subject to per-restaurant hourly order caps. Rotating session identifiers, creating multiple OAuth clients, or any other technique to circumvent per-identity limits is prohibited and may result in permanent account suspension.
(c) No prompt injection. You may not embed instructions in menu item names, descriptions, or any operator-controlled field that are designed to override, confuse, or redirect an AI agent acting on behalf of a customer.
(d) No fraudulent orders. Using automated agents to place, cancel, or modify orders for the purpose of disruption, financial gain at another's expense, or testing without authorization is prohibited.
(e) Anomaly-based enforcement. Our fraud gate automatically shadowbans OAuth identities that trigger repeated anomaly signals (rapid orders, duplicate carts, repeat refunds, scope violations). Shadowbanned agents receive structured error responses; appeals may be submitted to contact@synaltix.io.

You may only create a Thmenu account if you represent a genuine restaurant, café or other food-and-beverage establishment intending to serve real customers.

The following account patterns are prohibited and will result in immediate suspension without refund:

• No-intent accounts: creating an account with no genuine intention to operate a menu or accept customer orders
• Bait content: publishing a menu solely to drive traffic elsewhere, collect customer data under false pretences, or test competitor pricing without disclosure
• Impersonation: creating an account that impersonates an existing restaurant, brand, or public figure
• Synthetic reviews or feedback: submitting, soliciting, or artificially inflating customer feedback scores or NPS ratings

We may remove accounts that, after 90 or more days with 50 or more published products and zero genuine customer orders, show no credible sign of genuine restaurant operation. We will attempt to notify you before removal.

This section applies from 27 November 2026 (30 days' notice under §14).

Some panel features use AI to suggest content — for example menu text and translations, product details, allergen and nutrition estimates, images and replies to reviews. Every suggestion is a draft: it is shown to you before it is used, you decide whether to publish it, and you remain responsible for what your menu says, in particular allergens, ingredients, prices and legal notices. AI output can be wrong; check it against your recipes and supplier information.

Where it runs. Thmenu's AI features run on Cloudflare Workers AI; the text or image you submit is processed to produce the suggestion and Thmenu does not use it to train models.

What you may not do. You may not use AI features to generate content that infringes others' rights, impersonates another business, or misleads guests (for example invented reviews, awards or certifications). The guest-facing menu assistant answers from your menu data only and does not place orders or take payments by itself — the guest confirms.