fixing-metadata
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill identifies and audits metadata from external HTML pages, creating an ingestion point for potentially untrusted data. While this creates a vulnerability surface, the skill includes a best-practice instruction to escape and sanitize dynamic strings. No explicit boundary markers for external data are defined, and the skill's capabilities are focused on specific metadata file modifications.
Audit Metadata