specialist-product-behavioral-nudge-engine

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the interpolation of the {the_user_question} variable.
  • Ingestion points: User input is ingested via the {the_user_question} placeholder in the SKILL.md file within the Agent tool call.
  • Boundary markers: The skill lacks delimiters (e.g., XML tags or triple quotes) or specific 'ignore instructions within input' warnings to separate the system instructions from the user-provided content.
  • Capability inventory: The spawned sub-agent is designed to suggest actions, draft replies, and interact with user preferences. If the user input contains malicious instructions, the sub-agent could be coerced into providing biased or unauthorized advice.
  • Sanitization: There is no evidence of sanitization or validation of the user input before it is passed to the sub-agent prompt.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 08:50 PM
Security Audit — agent-trust-hub — specialist-product-behavioral-nudge-engine