Security audit
Security checks for vulnerabilities and agentic risk
This is a simple email-reply/template guidance skill with no executable code, persistence, credential use, or hidden data access.
Review the scope before installing: this skill is low risk, but its activation wording is broad, so use it when you specifically want email or communication-drafting help and avoid pasting secrets or unnecessary personal information.
The activation guidance is broad enough to match many generic communication requests, which can cause the skill to trigger outside its intended narrow use case. Overbroad routing increases the chance that users provide unnecessary email content or sensitive context to a generic assistant workflow, even though the skill itself includes otherwise safe guardrails.
The example prompt is phrased so generically that it overlaps with ordinary requests for planning help, making accidental activation more likely. In an agent ecosystem, vague triggers can misroute tasks to the wrong skill and weaken safety-by-specialization, especially when users did not intend to invoke an email-focused workflow.
No suspicious patterns detected.
