matlab-design-radar-waveform
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No attempts to override agent behavior, bypass safety guardrails, or extract system instructions were found. The instructional flow is strictly focused on technical domain-specific requirements gathering.
- [DATA_EXFILTRATION]: No unauthorized file system access or network exfiltration patterns were detected. All described operations utilize local MATLAB toolbox functions for signal generation and analysis.
- [REMOTE_CODE_EXECUTION]: The skill does not download or execute remote scripts from unverified sources. It relies on standard, versioned MATLAB toolbox functionality.
- [COMMAND_EXECUTION]: No evidence of arbitrary shell command execution or privilege escalation. The skill generates MATLAB script snippets for waveform design, which is the intended primary functionality.
- [OBFUSCATION]: All files were scanned for hidden payloads, encoding (Base64/Hex), zero-width characters, and homoglyphs; no such techniques were identified.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or credentials found. The skill follows best practices for parameter derivation from user requirements.
- [DYNAMIC_EXECUTION]: While the skill generates MATLAB code at runtime, this behavior is standard for its purpose and uses well-defined System Object templates without incorporating untrusted executable logic.
Audit Metadata