Vector Search Designer

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • SAFE (SAFE): The skill is strictly instructional. It provides documentation, tables, and pseudo-code snippets for educational and design purposes.
  • PROMPT_INJECTION (SAFE): No malicious instructions or attempts to bypass agent safety filters were detected in the text or metadata.
  • DATA_EXPOSURE (SAFE): There are no instances of hardcoded credentials, access to sensitive local file paths, or external network requests.
  • REMOTE_CODE_EXECUTION (SAFE): The skill does not perform any package installations (pip, npm) or download remote scripts.
  • COMMAND_EXECUTION (SAFE): No system commands or subprocess spawning are used; all code blocks are provided as non-executable examples for the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:19 PM