what-if-oracle

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists exclusively of markdown instructions and reference templates. It does not contain any executable scripts, binary files, or configuration files that could pose a security risk.
  • [PROMPT_INJECTION]: No evidence of prompt injection or safety bypass attempts was found. While the skill defines a persona ('The Oracle'), it does so for task framing without attempting to override system constraints or ignore safety protocols.
  • [EXTERNAL_DOWNLOADS]: The skill references external research using Zenodo DOIs and a GitHub repository for its upstream source. These are used as informational references and do not involve the automated downloading or execution of remote content.
  • [DATA_EXFILTRATION]: There are no commands or instructions that access sensitive files, environment variables, or perform network requests that could lead to data exfiltration.
  • [COMMAND_EXECUTION]: No shell commands, subprocess calls, or privileged operations are present in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 08:08 PM
Security Audit — agent-trust-hub — what-if-oracle