occlusion-trace-meta-proof

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: LOWNO_CODE
Full Analysis
  • NO_CODE (SAFE): The skill lacks any executable code, shell scripts, or external dependencies. It relies entirely on natural language instructions for the AI agent.
  • PROMPT_INJECTION (SAFE): No malicious instructions designed to bypass safety guardrails or hijack the agent's persona were identified.
  • DATA_EXFILTRATION (SAFE): The skill does not contain any commands to access sensitive files or transmit data to external servers.
  • EXTERNAL_DOWNLOADS (SAFE): No remote resources or third-party packages are referenced or downloaded.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 17, 2026, 12:25 PM