afk-advanced-elicitation

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected. The skill functions as an instruction-based framework for text refinement and critique.\n- [EXTERNAL_DOWNLOADS]: The skill references a local file ./methods.csv as its primary resource. No external network requests, URL fetches, or remote downloads are performed.\n- [COMMAND_EXECUTION]: No shell commands, subprocess calls, or system-level execution patterns are present in the instructions or the bundled resource.\n- [DATA_EXFILTRATION]: There are no network operations or patterns suggesting the transmission of user data or artifact content to external domains.\n- [PROMPT_INJECTION]: The skill includes explicit operating rules to prevent the assistant from assuming external personas or orchestration layers, and focuses strictly on the artifact under review. No malicious override patterns were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 09:41 PM
Security Audit — agent-trust-hub — afk-advanced-elicitation