seo-crawl-render
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: Ingestion points: Untrusted data enters the agent context via WebFetch for URLs and Read/Glob for local files as described in the Acquisition section of SKILL.md.
- [INDIRECT_PROMPT_INJECTION]: Boundary markers: The skill does not define delimiters or provide instructions to ignore embedded commands within the ingested content, leaving it vulnerable to data-driven instruction overrides.
- [INDIRECT_PROMPT_INJECTION]: Capability inventory: The skill utilizes high-privilege tools including Bash, Read, Glob, and WebFetch (listed in the frontmatter), which could be leveraged if the agent inadvertently follows instructions contained within crawled data.
- [INDIRECT_PROMPT_INJECTION]: Sanitization: There is no evidence of sanitization, escaping, or validation of the content retrieved from external sources before it is interpreted by the agent.
Audit Metadata