Compliance linter for wellness marketing copy

The idea

Supplement and wellness brands live one sentence away from an FDA warning letter: say a product treats, cures, or prevents anything and it’s an unapproved drug claim. The people writing the copy are marketers, not regulatory lawyers, and the line between an illegal disease claim and a legal structure/function claim is genuinely subtle. Today the options are an expensive compliance review or hoping for the best.

Generic writing checkers proved the inline-linter form factor — people pay $5.47/mo for LanguageTool — but grammar is a commodity and compliance is not. The wedge is a domain rulebook: FDA and FTC warning letters are public, they encode exactly which phrasings draw enforcement, and an LLM primed with those patterns can flag risk and offer a compliant rewrite in the same underline-and-hover gesture marketers already understand. Why now: this classification was too nuanced for wordlists alone, and LLMs made it buildable in a week.

Who pays

DTC supplement, skincare, and functional-food brands — specifically the marketing lead, who fears the warning letter personally. This is compliance-risk pricing, not grammar pricing: $49/mo per brand, $99/mo for agencies running multiple brands. Distribution: DTC and e-commerce communities, Klaviyo and Shopify ecosystems, and content around recent FDA warning letters, which brands actively search after each enforcement wave.

MVP scope

• MV3 extension linting text fields on Shopify, Klaviyo, and Meta Ads Manager
• Two-stage detection: hard-trigger wordlist plus Claude claim analysis
• Hover cards with risk tier, explanation, and one-click compliant rewrites
• Full-page scan producing a shareable risk report
• Missing-disclaimer detection on structure/function claims
• Skip for v1: Google Docs support, EU/UK claim regimes, image/ad-creative scanning, approval workflows

Build ClaimCheck, a Chrome extension that lints supplement and wellness marketing copy for FDA/FTC compliance risk as marketers type. Stack: Chrome extension (Manifest V3, content script + service worker), Next.js backend for auth, billing, and the API proxy, Claude API for claim analysis.

Core flow: 1) Marketer installs the extension and signs in; the content script activates on Shopify admin, Klaviyo, Meta Ads Manager, and any site the user whitelists. 2) As they type in text fields, debounced text is sent to the backend, which runs a two-stage check: a fast local regex/wordlist pass for hard-trigger terms (cures, treats, prevents, disease names, anti-inflammatory), then Claude with a system prompt encoding structure/function-claim rules and patterns from published FDA and FTC warning letters. 3) Risky phrases get a red or amber underline; hovering shows why it's risky, the risk tier, and 2 compliant rewrites (e.g. supports a healthy immune system instead of a disease claim) applied with one click. 4) A page-level scan button audits an entire live product page or email draft and produces a shareable risk report.

Pages (web app): dashboard with scan history, brand settings including product category, team seats, billing. Non-obvious details: never send keystrokes for non-whitelisted domains; cache verdicts by text hash so repeated scans are free and fast. Include the standard FDA disclaimer checker — flag pages making structure/function claims without it. Pricing: 20 free scans, then $49/mo per brand, $99/mo for agencies with 5 brands.