BanProof AI

What BanProof AI MCP server does

BanProof AI MCP server checks video scripts for language that may conflict with TikTok Shop content policies, the Amazon Associates Operating Agreement, or FTC disclosure guidance. It is intended for creators, affiliate marketers, and teams reviewing promotional scripts before recording or publishing.

The audit covers several categories of potential problems, including medical claims, guarantees, unsupported certifications, scarcity language, fake social proof, income claims, missing advertising disclosures, unsupported price comparisons, AI-generated content disclosures, and testimonial or before-and-after claims that need disclaimers. When a product URL is supplied, the review can also compare script claims with the product page.

How it works

The hosted MCP endpoint is available at https://banproof.io/mcp. An MCP-compatible client sends a script or violation notice to the relevant tool and receives structured review output or drafted text. Authentication uses OAuth 2.0 through a BanProof account.

audit_script accepts the complete script as a required script string. The optional product_url provides a product page for claim cross-checking. Results include flagged phrases, explanations, suggested rewrites, an overall risk level of safe, elevated, or high, and BanProof risk scores for individual violations and the script overall. These scores are BanProof's own severity rating rather than a platform-issued score.

generate_appeal takes the full violation notice and a platform value of tiktok, amazon, or youtube. It returns appeal text ready to paste into the relevant form, along with a character count compared with the platform limit. The README specifically identifies an 800-character TikTok limit.

Setup and configuration

Connect an MCP client to the hosted endpoint and complete OAuth 2.0 authentication. A free account at BanProof AI provides three credits per month without requiring a credit card. The documented compatible clients include Claude, Cursor, and other MCP-compatible clients.

No environment variables or local service configuration are documented for this hosted connection. The listed credit costs are one credit per script audit and two credits per appeal. Paid Creator and Brand plans provide larger monthly credit allocations; the Brand plan also lists five seats.

Tools and capabilities

• Scan a full creator script before recording.
• Identify exact phrases associated with policy risks.
• Suggest alternative wording for flagged claims.
• Check promotional claims against a supplied product page.
• Review FTC disclosure omissions such as missing #ad or #sponsored markers.
• Draft platform appeal text from a violation notice.

Limitations and notes

BanProof AI MCP server provides policy-oriented checks and rewrites, not an official platform decision or guarantee that content will be approved. Its risk score is not issued by TikTok, Amazon, or another platform. Product-page comparison depends on supplying product_url, and the appeal output is based on the notice text provided by the user.

The README lists detection for an AI voice LIVE ban scheduled for July 2026 and a TikTok policy requirement for AIGC or AI voice disclosure. Platform rules can change, so users should treat the results as a pre-publication review rather than a substitute for checking current platform policies.