analyze-accent-report
Pass
Audited by Gen Agent Trust Hub on Jun 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access attempts were detected. The skill is designed for analytical tasks using vendor-specific data.
- [COMMAND_EXECUTION]: The skill instructions mention using the Coval CLI or API to inspect runs and retrieve metrics. This represents legitimate tool usage associated with the skill's primary purpose and the vendor's (coval-ai) infrastructure.
- [PROMPT_INJECTION]: The skill processes untrusted external data such as transcripts and report summaries, which constitutes a potential surface for indirect prompt injection. However, the skill includes explicit guardrails to prevent the exposure of secrets or private data and does not utilize high-privilege capabilities that would enable a harmful exploit.
Audit Metadata