AuditSocials Compliance MCP server

AuditSocials Compliance MCP — Pre‑Publish Ad & Social Policy Check for AI Agents

A Model Context Protocol (MCP) server that checks AI‑generated social media and ad content against the current advertising and community policies of 8 platforms — Meta (Facebook/Instagram), TikTok, LinkedIn, Google Ads, YouTube, X (Twitter), Snapchat and Pinterest — before it is published. It returns the exact risky phrase, why it's risky, and a compliant rewrite, so posts, ad accounts and campaigns aren't flagged, demonetized, shadow‑limited or banned.

Give your AI writing assistant or agent a pre‑publish compliance guardrail: one self‑describing tool it can call on any draft — no prompt engineering, no SDK.

• 🛡️ 8 platforms, current policy — advertising + community guidelines for Meta, TikTok, LinkedIn, Google Ads, YouTube, X, Snapchat, Pinterest.
• 🎯 Actionable findings — the exact risky phrase, the severity, a confidence label, why it violates policy, and a compliant rewrite.
• 🤖 Built for agents — a single MCP tool with a rich description; your agent discovers and calls it automatically.
• ⚡ 30‑second setup — npx, drop in a free key, done. Works with Claude, Cursor, VS Code, Windsurf, Cline and any MCP client.
• 🆓 Free tier — 50 checks / month, no credit card.

Table of contents

• Who is this for
• What it catches
• Quickstart
• Client setup (Claude, Cursor, VS Code, Windsurf, Cline…)
• Usage — just ask your agent
• The tool
• Example response
• Pricing & limits
• How it works
• FAQ
• Links

Who is this for

• AI content & ad‑copy tools that generate social posts, captions, tweets, video scripts or ad creative and need a moderation / brand‑safety pass before publishing.
• Agentic workflows & autopilots (Claude, Cursor, custom agents) that draft and schedule content and must not get the account restricted.
• Marketing & growth teams and agencies running paid social across many platforms who want to catch policy risk before an ad is rejected or an account is flagged.
• Developers building social schedulers, DTC marketing SaaS, influencer tooling, or compliance features who want a policy check as an API/MCP call instead of maintaining 8 platforms' rulebooks themselves.

What it catches

The guardrail combines deterministic pattern rules with AI reasoning over live platform policy. It surfaces risk across categories such as:

• Health, wellness & supplements — unrealistic weight‑loss claims, "miracle" / "cure" language, unverifiable "clinically proven" endorsements.
• Finance & crypto — "guaranteed returns", "risk‑free", "get rich quick", "zero risk", unsubstantiated return figures.
• Misleading & unsubstantiated claims — superlatives, false urgency, implied outcomes you can't back up.
• Prohibited & restricted content — categories platforms disallow or gate (regulated goods, sensitive attributes, etc.).
• Platform‑specific advertising rules — the things that quietly get ads rejected or distribution reduced rather than outright banned.

Shortened. The full README is on GitHub.