problem-solving

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOWNO_CODE
Full Analysis
  • NO_CODE (SAFE): The skill is comprised entirely of markdown reference files used for agent reasoning and does not include any Python scripts, Node.js packages, or shell commands.
  • EXTERNAL_DOWNLOADS (SAFE): The attribution file references the Microsoft Amplifier project on GitHub. This is a trusted source per [TRUST-SCOPE-RULE], and the skill does not perform any automated downloading or execution of the remote content.
  • PROMPT_INJECTION (SAFE): No malicious instruction overrides, system prompt extraction attempts, or safety bypass patterns were identified within the instructional text.
  • DATA_EXFILTRATION (SAFE): No network-capable code or sensitive file path access patterns are present in the skill package.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 06:28 AM