Redaxa for AI Security

Review possible secrets and personal data before sharing AI prompts

Redaxa helps you review text for potential secrets and personal information before sharing it with an AI assistant. Paste a prompt, inspect the findings and review a redacted version before deciding what to send.

The scanner detects supported credential patterns, email addresses and selected financial and personal-data formats. Custom protected terms help you flag project names or other sensitive words specific to your work. Pattern and checksum checks reduce some false positives, but they cannot identify every secret or prove that every match is sensitive.

Prompt scanning uses Redaxa’s backend; it is not entirely on-device. Scanning does not send the text to an AI provider. See the official privacy policy for processing and retention details. An account and an active trial or subscription are required for protected scanning features. The browser extension is a separate installation, not included in this Windows installer.

The Windows installer is signed by Aurelio Avila and timestamped. Code signing identifies the publisher and protects file integrity; it does not guarantee that Windows SmartScreen will never display a warning.

Redaxa is a review aid, not an antivirus or a guarantee against data disclosure. Review every result before sharing sensitive content.

Pro and Business plans also support local GitHub repository checks for candidate secrets. Findings need manual review: Redaxa does not test credentials or establish that a match is valid or belongs to a particular account.