pii-scrub

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses local Python processing to identify and redact sensitive data based on a comprehensive list of standard regex patterns for common secrets (AWS, Google, GitHub, Slack, etc.).
  • [SAFE]: No network activity was detected; the script relies entirely on standard Python libraries (re, sys, collections) and does not perform any external downloads or data exfiltration.
  • [SAFE]: The tool follows safety best practices by writing to a new .redacted file by default and providing a summary of redacted items to the error stream for user review.
  • [SAFE]: Metadata and instructions are consistent with the implementation, focusing purely on data sanitization and privacy.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 10:00 PM
Security Audit — agent-trust-hub — pii-scrub