seo
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill implements an indirect prompt injection surface by instructing the agent to use the WebFetch tool to ingest content from external, untrusted websites during keyword research and competitor analysis (defined in SKILL.md sections 3.9 and 3.10). An attacker could place malicious instructions on a website analyzed by the agent to influence the agent's behavior during the 'Fix' phase.
- Ingestion points: Data enters the agent's context from external competitor sites and search results via WebFetch as described in SKILL.md (Modules 3.9, 3.10, 3.11, 3.14, and 3.19).
- Boundary markers: The skill does not provide explicit boundary markers or instructions to treat external web content as untrusted data or to ignore embedded instructions found within fetched pages.
- Capability inventory: The skill possesses file-write capabilities for applying SEO fixes and can execute shell commands (e.g., grep, npm build) for framework detection and verification.
- Sanitization: No specific sanitization or validation of the data retrieved from external websites is implemented in the skill's logic.
Audit Metadata