Security audit
Security checks for vulnerabilities and agentic risk
The skill’s research purpose is coherent, but it asks the agent to download or update and then run mutable external code from a repository outside the reviewed package.
Install only if you are comfortable with the agent fetching and updating sensedeal/cue-skills outside the reviewed skill package and running its Python runner with access to your Cue API key and credits. Prefer using an already reviewed local checkout or a pinned commit, and confirm credit-spending prompts carefully.
The natural-language instructions and user-facing description are entirely in Chinese across the file, with no indication that other languages are supported or that the user can choose their preferred language. This can violate a language/locale policy when a skill forces a specific language without user opt-in.
The skill instructs the agent to clone or update an external repository from GitHub or Gitee and then execute a Python runner from that codebase. This creates a software supply-chain and arbitrary code execution risk that is much broader than the declared research task, because repository contents can change over time and are not pinned, reviewed, or sandboxed. The skill context makes this more dangerous because the external code is presented as a prerequisite step for normal operation, increasing the chance an agent will fetch and run untrusted code automatically.
