product-marketing
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill file is composed entirely of markdown instructions and strategic templates. It does not contain any scripts, binaries, or automated tool configurations.
- [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were identified. The skill serves as a guide for a marketing persona.
- [INDIRECT_PROMPT_INJECTION]: The skill defines ingestion points for untrusted data including product features and competitive landscape data in SKILL.md. However, it is categorized as safe because the agent lacks any capabilities for command execution, network access, or file system modifications (Capability inventory: none). No boundary markers or sanitization methods are specified as they are not required for this purely instructional skill.
Audit Metadata