Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed LinkPix/qhkit image-generation workflow, with ordinary dependency and account setup risks but no evidence of hidden or malicious behavior.
Before installing, confirm you are comfortable using the qhkit/LinkPix service, uploading garment/model images to it, and providing its API token. Prefer a normal vetted qhkit installation over ad-hoc npx use, and approve any Node or PATH changes explicitly if your environment lacks Node.
The manifest description says the skill 'must trigger' whenever a user asks for model try-on, outfit display, or flat-lay-to-model effects, and then lists broad keywords such as '换装' and '上身图'. These phrases are common enough in general e-commerce/image-editing requests that they can overlap with adjacent skills, and the file does not provide negative examples or clear non-trigger conditions.
The skill recommends falling back to npx @iqinghu/qhkit <命令> ... without pinning a specific package version. That causes runtime retrieval of whatever package version is current at execution time, creating supply-chain risk and making behavior non-reproducible if the upstream package is compromised or a breaking/malicious release is published.
The skill directs the agent to install a Node binary into $HOME/.local/lib and modify PATH, which introduces persistent executable changes in the user environment outside the task's immediate scope. Even though the download includes checksum verification, the behavior still alters the session/user environment and can outlive the current task, increasing attack surface and violating least surprise.
cd /tmp && curl -fsSLO https://nodejs.org/dist/v22.22.3/node-v22.22.3-linux-x64.tar.xz cd /tmp && curl -fsSL https://nodejs.org/dist/v22.22.3/SHASUMS256.txt | grep ' node-v22.22.3-linux-x64.tar.xz$' | sha256sum -c - mkdir -p "$HOME/.local/lib" && tar -xJf /tmp/node-v22.22.3-linux-x64.tar.xz -C "$HOME/.local/lib" export PATH="$HOME/.local/lib/node-v22.22.3-linux-x64/bin:$PATH"
The skill directs the agent to install a Node binary into $HOME/.local/lib and modify PATH, which introduces persistent executable changes in the user environment outside the task's immediate scope. Even though the download includes checksum verification, the behavior still alters the session/user environment and can outlive the current task, increasing attack surface and violating least surprise.
cd /tmp && curl -fsSLO https://nodejs.org/dist/v22.22.3/node-v22.22.3-linux-x64.tar.xz cd /tmp && curl -fsSL https://nodejs.org/dist/v22.22.3/SHASUMS256.txt | grep ' node-v22.22.3-linux-x64.tar.xz$' | sha256sum -c - mkdir -p "$HOME/.local/lib" && tar -xJf /tmp/node-v22.22.3-linux-x64.tar.xz -C "$HOME/.local/lib" export PATH="$HOME/.local/lib/node-v22.22.3-linux-x64/bin:$PATH"
The instructions force use of the domestic Chinese site/account flow and explicitly tell the agent to steer users away from the international site unless they insist. This can misroute users to a different regional service, causing credential handling, privacy, compliance, or data-sovereignty issues when users did not knowingly opt into that region-specific platform.
The '何时触发' section gives sample utterances, but it does not define when the skill should not trigger or how to distinguish this skill from neighboring image-editing skills. Without explicit boundaries, routine requests about changing model style or clothing visuals may invoke this skill unintentionally.