Redact FAQs

Question 1

Does Redact offer features for compliance and auditing?

Accepted Answer

Absolutely. When enabled, Redact provides detailed compliance audit logging. It records every obfuscation and deobfuscation operation, including the full input text, all detected sensitive values with their replacements, and the resulting output text, creating a verifiable audit trail.

Question 2

What is Redact and how does it protect sensitive data?

Accepted Answer

Redact is a security tool that automatically detects and replaces sensitive data (PII, secrets, etc.) in text and HTTP requests with fake values. This prevents real client data from being exposed to AI models like Claude during analysis, enhancing data privacy and security.

Question 3

How does Redact ensure secure interaction with AI models like Claude Code?

Accepted Answer

Redact integrates as an MCP server and plugin for Claude Code. It includes an HTTP proxy tool that deobfuscates requests for real calls and then obfuscates responses before Claude sees them. Hooks provide leak detection, while slash commands enable manual mapping and export.

Question 4

Can the original data be restored after obfuscation?

Accepted Answer

Yes. Redact maintains a bidirectional mapping table, ensuring that every sensitive value replaced with a fake one can be losslessly restored to its original state. This is crucial for generating final reports for clients without exposing raw data to AI.

Question 5

How does Redact detect sensitive information?

Accepted Answer

Redact employs a dual-powered detection system: 19 types of regex pattern matching for structured data (e.g., IPs, emails, API keys) and AI-powered Named Entity Recognition (NER) for unstructured entities like person names, organizations, and locations.

Redact

Redact

Características Principales

Casos de Uso

Características Principales

Casos de Uso