viral-hook
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection due to the way it handles untrusted inputs and external data references.
- Ingestion points: Untrusted data enters the agent context through the
$ARGUMENTSparameter inSKILL.mdand via the content of thebrand-voice.mdfile. - Boundary markers: The instructions do not employ delimiters (such as XML tags or specific block markers) or provide instructions to the agent to disregard potential commands embedded within the external data.
- Capability inventory: The skill is configured to read several files from the local environment, including
viral-hook/meta.json,core/global-constraints.json,core/marketing-triggers.json, andbrand-voice.md. - Sanitization: There is no evidence of validation, escaping, or filtering of user-supplied arguments or file contents before they are incorporated into the prompt.
Audit Metadata