objection-destroyer
Fail
Audited by Snyk on Feb 20, 2026
Risk Level: CRITICAL
Full Analysis
CRITICAL E004: Prompt injection detected in skill instructions.
- Potential prompt injection detected (high risk: 0.80). The prompt includes repeated "disable-model-invocation: true" directives that attempt to change the agent's behavior (disabling model invocation) which is unrelated to the skill's stated purpose of crafting pitch closings and thus constitutes a deceptive prompt-injection instruction.
Audit Metadata