react-effects-audit

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No prompt injection patterns were found. The instructions focus on a specific workflow for auditing code without attempting to bypass safety constraints.
  • [DATA_EXFILTRATION]: The skill does not contain instructions to access sensitive files or perform network exfiltration. It only reads source code files for analysis.
  • [REMOTE_CODE_EXECUTION]: There is no evidence of remote script execution, untrusted package installation, or dynamic code evaluation.
  • [COMMAND_EXECUTION]: The skill mentions 'Grep' for pattern matching, which is used as an internal processing step for the agent to analyze file content and does not involve unsafe shell command execution.
  • [SAFE]: The skill's functionality is transparent, well-documented, and strictly limited to its stated purpose of improving code quality.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 11:02 AM