bbq-and-smoking-expert

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for culinary education and utilizes a modular structure for research and validation. All files contain standard instructional content without security risks.
  • [PROMPT_INJECTION]: The skill defines a research checklist that involves ingesting external data, which is an indirect prompt injection surface. 1. Ingestion points: The skill instructs the agent to use adn_skills to distill content from Wikipedia and Arxiv in modules/research-checklist.md. 2. Boundary markers: The checklist includes explicit steps to validate legacy guidance against new sources, acting as a functional boundary. 3. Capability inventory: The skill focuses on culinary knowledge and does not include capabilities for file system modification, network exfiltration, or code execution. 4. Sanitization: No explicit technical sanitization is described, relying on the agent's internal safety protocols for processing tool outputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 05:38 PM
Security Audit — agent-trust-hub — bbq-and-smoking-expert