Prompt Leak Guard

Prompt Leak Guard helps catch accidental secret leaks before you paste real work into AI tools. It scans prompts, logs, configs, HAR exports, browser storage dumps, and debug snippets with 80+ local patterns for AI provider keys, cloud credentials, developer/package tokens, webhook URLs, signed URLs, auth headers/cookies, browser storage tokens, private keys, credential-bearing database and broker URLs, payment cards, and optional personal data like emails, phone numbers, and dashed US SSNs.

v0.1.5 adds browser-context leak detection for HAR authorization headers, HAR cookie headers, localStorage/sessionStorage token writes, and browser storage dump token fields. It keeps the existing risk receipts: each scan explains what matched, why it matters, the recommended next action, and a safer prompt hint.

Free browser scanner: try the local scanner at https://site-mocha-three-50.vercel.app/ai-prompt-secret-scanner before buying.

Launch pricing: list price is $4.99. Use code LAUNCH20 for 20% off the first 100 purchases.

The scanner and license check run locally in the browser. There is no backend, no analytics SDK, and no remote model call. The download includes the v0.1.5 browser extension ZIP, launch license code, install notes, privacy/support docs, QA report, changelog, and AI prompt safety checklist.

Best for developers, analysts, operators, security researchers, and AI-heavy teams who regularly paste logs, configs, HAR snippets, browser context, prompts, or debug output into AI chats. This is a pattern-based safety check, not a guarantee. Rotate any credential that may already have been exposed.