Security audit
Security checks for vulnerabilities and agentic risk
The skill is a CSV-to-JSON helper, but it asks for broad read/write/command authority and describes API and command execution behavior beyond that narrow purpose.
Install only if you are comfortable with a CSV conversion skill that may be invoked for broader automation tasks and may use read, write, command execution, API keys, and external integrations. Prefer running it in a restricted workspace, avoid giving it sensitive CSVs or credentials unless necessary, and review any command or outbound API action before approval.
The activation text is overly broad, covering generic API integration, webhook setup, system connectivity, and automation scenarios far beyond CSV-to-JSON conversion. In agent environments, broad matching can cause the skill to trigger in unrelated contexts, exposing exec/write/network capabilities where the user expected a simple data transformation.
The description explicitly states '支持中文交互' as a behavioral expectation, and the display metadata is localized to Chinese without offering a user language choice. This can violate language/locale policy when the skill appears to force or strongly bias a specific language without opt-in or justification.
The description claims risky code and external dependencies were removed, yet the same file still promotes exec, write, API key configuration, network communication, and external integrations. This creates a false sense of safety and may cause users or orchestrators to trust and invoke a skill with materially broader risk than disclosed.
The skill is presented as a narrow CSV-to-JSON utility, but later sections advertise broader capabilities including file handling, API integration, command execution, and information retrieval. This mismatch expands the effective authority of the skill beyond user expectations, increasing the risk of unintended system changes, data exfiltration, or unsafe execution when the skill is auto-selected or trusted based on its title.
The markdown describes file writing, API calls, and command execution as features but does not provide a clear, prominent warning that these actions can modify the filesystem, invoke external services, consume credentials, or affect the local environment. Users may approve or invoke the skill under the assumption it only converts CSV data, which raises the chance of unsafe side effects.
No suspicious patterns detected.
