lichtenstein-creative-testing
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill package is composed entirely of markdown documentation, templates, and metadata files. No malicious patterns, obfuscation, or unauthorized access attempts were detected during the analysis.
- [NO_CODE]: The skill does not contain any executable scripts (such as .js, .py, or .sh files) or compiled binaries. It operates purely as a set of instructions for the agent.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing untrusted user-supplied data such as product descriptions and creative briefs.
- Ingestion points: External data is ingested when the agent is instructed to read account facts, product details, and creator briefs as part of the planning procedure in
SKILL.md. - Boundary markers: There are no explicit delimiters or specific instructions to ignore embedded commands within the user-provided content in the templates.
- Capability inventory: The skill's capabilities are limited to text generation and plan formatting; it does not request or use high-privilege tools such as shell access, network operations, or file-system modification scripts.
- Sanitization: No input validation or sanitization mechanisms are implemented for the ingested marketing data.
Audit Metadata