skills/google/mantis/mantis-calibrate/Gen Agent Trust Hub

mantis-calibrate

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFE
Full Analysis
  • Automated Script Utility: The skill involves the creation and execution of a localized Python helper script (workspace/helpers/append_calibrate.py) to manage file updates. This method is an efficiency-oriented design choice to reduce data redundancy during the calibration process.
  • Data Analysis Surface: As a risk calibrator, the skill processes external findings, which introduces a data ingestion surface common to security tools.
  • Ingestion points: Findings JSON files in workspace/findings/ and the project THREAT_MODEL.md file.
  • Boundary markers: The implementation utilizes structured JSON formats and predefined enums to maintain data integrity.
  • Capability inventory: File system access (read/write) and execution of the generated helper script for JSON manipulation.
  • Sanitization: The skill applies field normalization and specific fallback defaults to ensure the scoring matrix remains consistent even with incomplete input data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 01:21 PM
Security Audit — agent-trust-hub — mantis-calibrate