anti-deception
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill depends on the external ejentum MCP server, which is a third-party resource relative to the author.
- [DATA_EXFILTRATION]: The skill transmits a summary of the user's interaction (the 'query' argument) to the external ejentum.com infrastructure via the anti-deception tool call.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted user input to identify deception patterns. While the skill is designed to improve agent integrity, it lacks explicit boundary markers or sanitization steps before using user-derived context to trigger external tool calls that influence the agent's internal reasoning and output strategy.
Audit Metadata