58-seo
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides architectural and implementation guidelines for technical SEO. No malicious patterns, such as prompt injection, data exfiltration, or persistence mechanisms, were detected.
- [NO_CODE]: The skill contains only markdown instructions and does not include any executable code, shell scripts, or external dependencies.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to inspect the existing repository and user requirements, which constitutes an ingestion surface for untrusted data. However, as the skill's capabilities are restricted to implementation guidance without dangerous tools (like network exfiltration or shell execution), this surface is considered safe.
- Ingestion points: Existing repository and user requirements (SKILL.md).
- Boundary markers: None specified.
- Capability inventory: Repository inspection and code implementation.
- Sanitization: None specified.
Audit Metadata