Secret detection
Flags common AI provider keys, AWS/Azure/GCP credentials, GitHub/GitLab/npm/PyPI tokens, chat/webhook URLs, payment/email/ops platform keys, bearer tokens, validated JWTs, private keys, and credential-bearing database URLs.
Local-only browser extension
Prompt Leak Guard scans AI prompts, debug logs, and config snippets for common API keys, cloud/service tokens, webhooks, private keys, signed URLs, credential-bearing database URLs, payment cards, email, phone, and dashed US SSN patterns before you paste them into AI tools.
Designed for developers, operators, analysts, and AI power users who paste real work into AI tools and need a fast local safety check first.
Flags common AI provider keys, AWS/Azure/GCP credentials, GitHub/GitLab/npm/PyPI tokens, chat/webhook URLs, payment/email/ops platform keys, bearer tokens, validated JWTs, private keys, and credential-bearing database URLs.
Replaces sensitive values with readable placeholders so the prompt keeps its shape without leaking the real value.
On supported AI chat pages, the extension warns when high-risk content appears in the active prompt field.
No account, no cloud scanner, no analytics SDK, and no remote model call. The detection engine runs in the browser and outputs sanitized text locally.
The extension is static JavaScript. There is no backend to maintain and no API bill to cover.
The launch ZIP includes the detector code, popup, options page, and manifest so buyers can inspect what is installed.
Paid access uses an offline license code. Activation is local, with no account system or server check.
Packaged extension ZIP, offline license code, local detector engine, popup scanner, supported-site warning script, and install notes.
Launch code LAUNCH299: 40% off for the first 50 buyers, about $2.99 before tax. Checkout is handled through Payhip. The download includes the buyer ZIP and a launch license code text file so activation works immediately after purchase.The launch build is intentionally small: one local utility, one job, no account system, no background service.
No. It catches common high-risk patterns and helps prevent obvious accidental leaks. It is a safety check, not a guarantee or a replacement for secret rotation.
A packaged browser extension ZIP, license code, install notes, privacy/support docs, and the auditable detector code included in the extension package.
In Chromium-style browsers through Manifest V3. The content warning is scoped to supported AI chat pages listed in the privacy policy.