persona-sim
Warn
Audited by Socket on Jun 27, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: The core capability is coherent with persona simulation, and installation is relatively standard, but the skill intentionally supports custom LLM gateways and third-party OpenAI-compatible endpoints that can receive prompts and API keys instead of official provider APIs. This is not fundamentally incompatible with the stated purpose, but it creates medium data-flow and credential-routing risk that should be treated cautiously.
Confidence: 86%Severity: 58%
Audit Metadata