security-compliance-privacy
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill configuration enforces strict security boundaries by explicitly forbidding the agent from reading sensitive file patterns such as .env files, secrets directories, and cryptographic keys (*.pem, *.key).- [SAFE]: Comprehensive redaction policies are established to prevent the exfiltration or exposure of real personal data (PII) in model outputs, specifically masking emails, names, and IP addresses.- [SAFE]: The skill explicitly forbids the use of web-fetching tools, significantly reducing the risk of unauthorized network communication or data exfiltration.- [INDIRECT_PROMPT_INJECTION]: The workflow for processing Data Subject Access Requests (DSAR) involves ingesting potentially untrusted input from external users, creating a known attack surface.
- Ingestion points: External user requests received via email, in-app forms, or web forms (SKILL.md).
- Boundary markers: The instructions do not define specific delimiters to isolate external request content from the system prompt.
- Capability inventory: The skill allows the use of Read, Grep, Glob, and Bash tools for data processing.
- Sanitization: The skill mandates identity verification and strict output redaction as primary defensive measures.
Audit Metadata