gcp-agent-tool-trajectory-evaluator

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of two files providing metadata and Python logic for evaluation metrics. Analysis of scripts/trajectory_metrics.py shows standard data processing for precision and recall calculations without any dangerous operations.
  • [SAFE]: No network requests, file system access, or credential usage were detected in the source code.
  • [SAFE]: No obfuscation or hidden instructions were found in the skill files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 11:49 AM