run-audio-quality-testing

Pass

Audited by Gen Agent Trust Hub on May 26, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the official coval CLI tool to interact with the vendor's platform. These commands are limited to listing resources (agents, test sets, metrics, personas) and launching testing runs, which aligns with the skill's stated purpose.
  • [EXTERNAL_DOWNLOADS]: The documentation provides instructions for installing the Coval CLI via Homebrew (brew install coval-ai/tap/coval). This is a standard and secure installation method for the vendor's official tooling.
  • [DATA_EXFILTRATION]: There is no evidence of sensitive data exfiltration. The skill handles platform-specific metadata like organization slugs and run IDs to construct report URLs on the official app.coval.dev domain.
  • [PROMPT_INJECTION]: The instructions do not contain any bypass attempts, role-play overrides, or system prompt extraction patterns. The logic focuses on structured automation of testing tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
May 26, 2026, 04:49 PM
Security Audit — agent-trust-hub — run-audio-quality-testing