ai-visibility-audit
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches external data using curl against user-provided URLs as defined in 'references/checks.md'. This creates a surface for indirect prompt injection. 1. Ingestion points: 'references/checks.md' (curl calls to robots.txt and HTML content). 2. Boundary markers: Absent. 3. Capability inventory: 'references/checks.md' (curl, python3 -m json.tool). 4. Sanitization: Absent. The risk is inherent to the auditing function and is mitigated by the use of specific grep filters to limit the data returned to the agent.\n- [REMOTE_CODE_EXECUTION]: The skill utilizes common shell utilities (curl, grep, sed) and a standard library Python module (json.tool) for processing data. These operations are restricted to the task of auditing public web content and do not present a risk of arbitrary code execution, privilege escalation, or unauthorized system access.
Audit Metadata