Security audit
Security checks for vulnerabilities and agentic risk
This career-tracking skill fits its job-search purpose, but it would persist sensitive career information too broadly and without clear user confirmation.
Install only if you are comfortable with the assistant maintaining a persistent job-application file and reading career-profile information. Treat career conversations as potentially state-changing unless the skill is revised to require confirmation before writing updates.
The skill description includes broad activation language such as general job-search assistance tasks without strong guardrails on when it should engage. In an assistant environment, that can cause unintended invocation during ordinary conversation, exposing personal career context or triggering follow-on actions the user did not explicitly request.
The natural-language update triggers are ambiguous shortcuts that map casual phrases like 'heard back' or 'got an offer' directly to workflow updates. This creates a risk of unintended skill activation or incorrect state changes when the user is speaking hypothetically, quoting someone else, or providing partial information.
The instruction to update the JSON immediately after any career conversation mandates persistent writes without user notice or consent boundaries. In practice, this can silently store sensitive employment information, create inaccurate records from exploratory chats, and normalize state changes without confirmation.
No suspicious patterns detected.
