matlab-analyze-ams-waveform

Pass

Audited by Gen Agent Trust Hub on Aug 2, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill identifies ingestion points for local data files including .mat, .csv, and HSpice simulation formats (.tr0, .ac0, .sw0) as seen in Phase 1.1 of SKILL.md. It also saves temporary plot files to the system's temporary directory for display purposes. These operations are standard for data analysis and do not involve exfiltration to external domains.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted waveform data from external files. While it lacks explicit boundary markers to delimit user-provided data from system instructions, the data is ingested using standard numerical parsers (readmatrix, load) and processed as signal vectors, which minimizes the attack surface for instruction injection.
  • [COMMAND_EXECUTION]: The skill uses the evaluate_matlab_code tool to perform signal processing calculations. The code snippets provided in the instructions are restricted to the Mixed-Signal Blockset API and standard MATLAB mathematical functions.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 2, 2026, 12:12 PM
Security Audit — agent-trust-hub — matlab-analyze-ams-waveform