pii-in-unstructured
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified. The skill adheres to its stated purpose of data privacy and classification.
- [INDIRECT_PROMPT_INJECTION]: The skill serves as a processing engine for untrusted unstructured data (emails, chat logs, documents). While this presents a surface for indirect prompt injection, the skill lacks exploitable capabilities to realize this threat.
- Ingestion points: The
scan_documentfunction inscripts/process.pyaccepts arbitrary string input from various source systems. - Boundary markers: No delimiters are currently used to wrap the input text.
- Capability inventory: The skill is restricted to regex matching and string manipulation; it contains no subprocess execution, network requests, or file-writing logic.
- Sanitization: The skill proactively masks detected PII (e.g., credit card numbers and NINOs) within the
value_maskedfield, ensuring sensitive data is not leaked into the scan results.
Audit Metadata