Security audit
Security checks for vulnerabilities and agentic risk
This sales research skill is not malicious, but it should be reviewed because it directs agents to find and infer individual contact details with broad activation and limited user confirmation.
Review this skill before installing if your environment has privacy, compliance, or outreach rules. It should be used only for user-directed B2B account research, with explicit approval before collecting or inferring individual emails or contact channels, and with shell/file operations constrained to generated report artifacts.
The skill description and activation criteria do not warn users that it will collect, infer, and surface personal contact information for identifiable individuals. This lack of transparency is dangerous because users may unintentionally trigger a workflow that performs people-search, contact enrichment, and targeting operations with significant privacy and compliance implications.
The trigger phrases are broad enough to activate on ordinary sales-related conversation, causing the skill to begin aggressive collection and analysis without strong user confirmation. In this skill's context, overbroad activation is more dangerous because the workflow proceeds into organization mining and contact-detail gathering, potentially exposing personal data when the user did not clearly request that level of profiling.
The skill metadata, triggers, examples, and operating instructions are entirely written for Chinese-language interaction, and line L023 directs immediate action without any indication that other user languages are supported. This creates a language-policy concern because the skill implicitly constrains interaction to a specific language without user opt-in or a documented regional justification.
The skill explicitly instructs the agent to discover individual email addresses and infer corporate email patterns for named people. That goes beyond high-level org mapping into targeted contact enrichment of personal data, which materially increases privacy, harassment, and unauthorized outreach risk when combined with role analysis and outreach recommendations.
The skill declares use of bash/file operations even though its stated purpose is sales research and visualization. Unnecessary shell capability expands the attack surface because adversarial content or future prompt changes could steer the agent into unintended local file access, command execution, or unsafe artifact handling.
No suspicious patterns detected.
