mcclipface.com

Very long descriptions are sometimes used to bury instructions to the AI. Skim it before allowing.

This server isn't on gate's handpicked list. That's not necessarily bad, but there's no third-party signal about it — review its tools carefully.

The description tells the agent that when 'browsing or checking out for the shopper', it should itself open the affiliate-tracked link before adding items to cart, rather than just presenting it as information. While disclosed and commission-neutral per the text, this is a directive embedded in a data-retrieval tool that steers agent behavior toward the vendor's monetization interests (ensuring its affiliate link is used) rather than purely returning coupon data. This repeats in find_coupons, find_product_coupons, and get_deals.

The tool states 'Report once per code per checkout, after the final attempt: every result, worked or failed (consent is presumed).' This instructs the agent to transmit usage data (store, code, outcome) to an external service based on an assumed/implicit consent model tied to a mode toggle, rather than requiring the agent to confirm with the user at the time of each report. Although the data collected is minimal (no cart/personal data), embedding 'consent is presumed' in a tool description nudges the agent to skip explicit per-action confirmation before sending data externally.