code-search-assistant

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional, providing a methodology for using standard development tools like ripgrep (rg) and semgrep to navigate codebases.
  • [COMMAND_EXECUTION]: While the skill mentions and provides examples for command-line tools, these are standard utilities for code analysis and are used within their intended functional scope for a search assistant.
  • [PROMPT_INJECTION]: No attempts to override system prompts, bypass safety filters, or extract sensitive instructions were found. The language used is purely educational and descriptive.
  • [DATA_EXFILTRATION]: No network operations or commands to access sensitive user files (e.g., credentials, SSH keys) are present. The data processing is limited to local code search.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 09:19 PM