elite-testing

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's functionality is limited to providing information and code templates for software testing. It does not include logic for network exfiltration, local file access, or unauthorized command execution.\n- [PROMPT_INJECTION]: While the skill contains text matching prompt injection signatures in llm-evals.md, these are used as examples to guide users in performing security evaluations of their own LLM applications and do not target the agent's own safety protocols.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 03:46 AM
Security Audit — agent-trust-hub — elite-testing