dry-three-copy-tests
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data, such as customer reviews, support notes, and interview transcripts, to generate marketing copy. * Ingestion points: The process is defined in SKILL.md (Step 3) and operationalized in references/method.md under the 'Build the evidence base' section. * Boundary markers: The skill employs structured templates in templates/output-template.md and uses specific identifiers (e.g., [E1], [S1]) to categorize and isolate different types of information. * Capability inventory: The skill does not utilize any scripts, shell commands, or external tools that could be exploited via data injection. * Sanitization: The instructions include explicit directives to separate attributable customer comments from factual product promises and to avoid inventing testimonials.
- [SAFE]: The skill consists entirely of text-based instructions and templates. It does not perform any network operations, file system modifications, or dynamic code execution. All external references are to well-known marketing educational resources for the purpose of attribution.
Audit Metadata