model-evaluator-validator

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No instructions were found that attempt to override agent behavior, bypass safety filters, or extract system prompts. The instructional tone is strictly professional and task-oriented.
  • [DATA_EXFILTRATION]: No evidence of hardcoded credentials, access to sensitive file paths (such as .ssh or .aws), or unauthorized network operations. The data handled is localized to prediction results and statistical outputs.
  • [REMOTE_CODE_EXECUTION]: The skill does not attempt to download or execute code from remote URLs or untrusted external sources.
  • [COMMAND_EXECUTION]: The workflow involves generating and executing Python scripts using well-known libraries like scipy.stats for model validation (McNemar test, T-Tests, etc.). This behavior is standard for the skill's primary purpose and does not involve risky command-line operations.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data (e.g., predictions.csv) as an ingestion point. While this represents a potential attack surface for indirect injection, the instructions focus on numerical and statistical processing which significantly limits the effectiveness of such attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 11:44 AM
Security Audit — agent-trust-hub — model-evaluator-validator