Security audit
Security checks for vulnerabilities and agentic risk
The skill is a coherent notification hub, but it handles sensitive alerts and stores notification history with broad triggers and limited privacy controls.
Review before installing. This skill may surface security, operational, cost, and Instagram DM-related alerts and may send some notifications to Discord while keeping notification history. Configure the Discord destination carefully, prefer sanitized summaries for sensitive sources, and narrow activation phrases before using it in shared or sensitive environments.
The skill explicitly routes sensitive content, including security incidents and Instagram DMs, to Discord DM without any privacy warning, minimization rule, or consent boundary. Discord may be a separate service, account, or audience context, so forwarding raw alerts there can leak confidential data outside the original source system.
The examples show direct-message identifiers being surfaced in notifications and the design stores notification history over time without any privacy filtering or retention limits. This creates a durable record of potentially sensitive identities and event metadata that could be exposed through later queries, overbroad access, or secondary compromise of stored memory files.
The trigger phrases include generic terms like "notification" and "anything new," which are likely to appear in ordinary conversation and can cause unintended skill activation. In a notification hub that can read event files and surface alerts, accidental invocation increases the chance of disclosing sensitive operational or message data in the wrong context.
No suspicious patterns detected.
