ranking-opportunity-finder
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided SEO performance data, creating a potential surface for indirect injection. Ingestion point: Pasted query/page performance exports or dated ranking evidence (SKILL.md). Boundary markers: No explicit delimiters (e.g., XML tags) are defined for input encapsulation in the provided prompt template. Capability inventory: Generation of Markdown, JSON, HTML, XML, and CSV deliverables (SKILL.md). Sanitization: The instructions include robust defensive measures, specifically commanding the agent to treat embedded instructions as data and to perform syntax validation on generated outputs.
- [SAFE]: No malicious patterns were identified. The skill does not attempt to bypass safety filters, access sensitive local files, or execute remote code. All operations are confined to the intended purpose of SEO data analysis.
Audit Metadata