iso27001-gap
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill incorporates explicit instructions for the AI agent to ignore any commands embedded in the documents it analyzes, treating them strictly as untrusted data for review.
- [SAFE]: Tool access is limited to basic file reading and searching (Read, Grep, Glob), which are necessary and appropriate for the skill's auditing function.
- [SAFE]: Static analysis signals regarding instruction overrides were found to be false positives; the identified patterns are defensive safety notices rather than malicious prompt injections.
- [SAFE]: Indirect prompt injection surface identified: 1. Ingestion points: Organizational documents read via Read, Grep, and Glob tools. 2. Boundary markers: Explicit 'Prompt Injection Safety Notice' provided in SKILL.md. 3. Capability inventory: File reading and pattern matching (Read, Grep, Glob). 4. Sanitization: Instructional guardrails telling the agent to treat content as data only. This surface is well-managed and inherent to the auditing purpose.
Audit Metadata