Security audit
Security checks for vulnerabilities and agentic risk
This is a markdown-only wealth-management workflow skill with sensitive financial-planning prompts, but no hidden execution, persistence, credential handling, or data exfiltration behavior.
Install only if you intend to use an agent for wealth-management preparation and are comfortable providing sensitive financial and household details. Ask the agent to confirm before looking up client records, producing retirement projections, or drafting rebalancing or tax-loss-harvesting recommendations.
The phrase "prep for [client name]" is overly generic and can be triggered by routine planning language unrelated to wealth-management workflows. Because this skill handles advisory and account-review functions, a false activation could pull sensitive client context or initiate financial analysis without clear user intent.
The phrase "prep for [client name]" is overly generic and can be triggered by routine planning language unrelated to wealth-management workflows. Because this skill handles advisory and account-review functions, a false activation could pull sensitive client context or initiate financial analysis without clear user intent.
The trigger phrase "can I retire" is common conversational language and may be used hypothetically or generally, not as a request to run a financial-planning workflow. In this domain, broad matching risks launching a planning skill that solicits or infers highly sensitive financial, health, and household data without sufficiently clear intent.
The file presents section headings and descriptive content in Chinese while command descriptions and workflow content remain in English, but it does not state whether the user can choose their preferred language. This can violate a language/locale policy when language is effectively imposed without opt-in.
