threat-modeling-ics
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill implements a comprehensive and structured workflow for ICS threat modeling, anchored in the STRIDE methodology and industry-standard security frameworks.
- [EXTERNAL_DOWNLOADS]: The skill references data sources and schemas from trusted organizations including MITRE and FIRST. These are used neutrally for threat enrichment and CVSS validation.
- [INDIRECT_PROMPT_INJECTION]: The skill provides explicit sanitization instructions to defend against indirect prompt injection or malformed input data. Section 4.1.1 (Sanitization) directs the agent to discard non-printable characters, Base64 strings, binary blobs, and image tags encountered in input fields.
- Ingestion points: Reads and processes Microsoft TMT export CSV files and model files (.tm7).
- Boundary markers: Establishes the CSV as the source of record and utilizes hygiene filters before extracting field values.
- Capability inventory: The skill is limited to reading local files and generating Markdown/CSV reports; it does not request network access or shell command execution.
- Sanitization: Implements an extensive hygiene policy targeting binary data, encoding mismatches, and non-semantic content to ensure the agent only acts on legitimate threat-modeling data.
Audit Metadata