Security audit
Security checks for vulnerabilities and agentic risk
This is a broad automation skill that is not malicious, but it asks for powerful file and shell access while under-specifying deletion, network/API, and scheduling boundaries.
Review this skill before installing if you plan to use it on important files or sensitive data. Use explicit task scopes, require previews and confirmations for cleanup or sync operations, avoid giving it API keys unless you know which service and data are involved, and do not enable scheduled runs unless you intentionally configure them.
The invocation guidance is so broad that the skill may be selected for many generic efficiency or workflow requests. In combination with Read/Write/Edit/Bash capabilities, overbroad routing increases the chance of unnecessary access to files or execution pathways when a safer, narrower skill would suffice.
The skill advertises duplicate-file cleanup and deletion-like behavior without a prominent warning about irreversible data loss. In an automation context, users may run bulk operations at scale, so even small logic errors or misunderstood filters can cause destructive loss of files.
Telling users to directly invoke the skill for any automation task without clear boundaries encourages misuse and over-selection. Because this skill advertises workflow execution and file operations, ambiguous invocation can expose local data or trigger unintended actions in contexts where a read-only or advisory skill would be safer.
The skill is presented as a local automation recipe pack, but later documentation expands scope to external API access, network use, and broader integrations. That mismatch can cause users or orchestration systems to grant or trigger the skill under false assumptions, increasing risk of unintended data transmission or actions outside the expected local-file context.
Documenting external API integration without clear necessity broadens the skill's effective privilege and can lead to sensitive data being sent to third-party services. In an automation skill handling files and structured data, this is dangerous because users may supply local content assuming it remains local.
The documentation mentions API integration and external service use without a clear notice that data may be transmitted off-host. For a skill centered on file and data processing, that omission creates privacy and compliance risk because users may provide sensitive local content without informed consent.
The documentation advertises system command execution beyond a narrowly defined recipe workflow. Since the manifest already includes Bash and the skill targets broad automation scenarios, vague command-execution capability can be abused for arbitrary local actions, including file modification, deletion, or environment inspection.
The display name and nearly all user-facing documentation are presented in Chinese, while the supported agent environments are otherwise broad and not region-specific. Because the document does not state that the skill is China-specific or offer a language choice, this can constitute an implicit language/locale constraint without user opt-in.
No suspicious patterns detected.
