ClauseGuard — Toxic Clause Checker for Terms

ClauseGuard helps you spot unfair or risky clauses in sign-up terms — Terms of Service, Privacy Policies, marketing-consent text, and more — **before** you agree. It is **BYOK (Bring Your Own Key)**: there is no bundled model and no server run by the developer. You connect an AI model you already use, and analysis runs directly between your browser and that endpoint. **Features** - One-click check of the current page's terms (or select text and check). - Detects multiple consent sections (ToS / privacy / marketing) and analyzes each. - Handles long terms via map-reduce chunking. - Findings sorted by severity, with plain-language explanations. - Click a finding to scroll to and highlight the original clause on the page. - Copy results as Markdown. - Optional URL-based suggestion badge for sign-up/terms pages (URL only). **Your privacy** - The developer operates no servers and collects no data — no analytics, no tracking. - Your API key is stored encrypted on your device (Web Crypto); an optional passphrase lock is available. - Terms text is sent **only** to the AI endpoint **you** configure, directly from your browser. That provider's policy governs that data. **Supported providers:** OpenAI, Anthropic (Claude), Google Gemini, OpenRouter, local models (Ollama / LM Studio), and any OpenAI-compatible custom endpoint. > ClauseGuard's analysis is for reference only and is not legal advice.