production-design-worldbuilding
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to perform a web research pass using external tools to gather details on cultures, technologies, and materials. This ingestion of untrusted data creates a surface for indirect prompt injection attacks where instructions hidden in web content could influence agent behavior.
- Ingestion points:
SKILL.mdworkflow explicitly requires a "small web research pass" before final output generation. - Boundary markers: No delimiters or specific instructions to ignore embedded commands within the research results are defined.
- Capability inventory: The skill is generative and utilizes external research tools; no sensitive file-system access or shell execution capabilities are defined within the skill itself.
- Sanitization: The instructions lack any requirement for the agent to sanitize or validate the content retrieved from the web.
Audit Metadata