Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed clothing recolor workflow that installs and uses the qhkit CLI to upload garment images for color-variant generation, with ordinary dependency and token-handling cautions.
Install only if you are comfortable adding the qhkit CLI and possibly Node to the user's environment, uploading selected clothing images to the Qinghu service, and configuring a qhkit API token that may spend account credits. Prefer a pinned package version or a vetted install path where possible, and avoid adding PATH changes to shell startup files unless the user explicitly wants that persistence.
The trigger description is very broad and includes common phrases such as recolor/change color/SKU color card, making accidental invocation more likely. Over-broad activation is a security concern because it can cause the agent to enter a tool-using workflow, install software, request credentials, or process files in conversations that only loosely match the intended task.
The skill recommends executing npx @iqinghu/qhkit without pinning a specific version. That causes each invocation to resolve code from the registry at runtime, which can unexpectedly pull a changed or compromised release and execute it in the agent environment. In this skill, that risk is amplified because the package is granted authority to process local files and handle API tokens.
This duplicated finding points to the same persistence behavior: unpacking software into a user-controlled directory and prepending it to PATH. That can influence future command resolution beyond the immediate task and is especially sensitive in an agent skill that also instructs downloading tools and handling authentication tokens.
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:$PATH"
This duplicated finding points to the same persistence behavior: unpacking software into a user-controlled directory and prepending it to PATH. That can influence future command resolution beyond the immediate task and is especially sensitive in an agent skill that also instructs downloading tools and handling authentication tokens.
cd /tmp && curl -fsSLO https://nodejs.org/dist/v22.22.3/node-v22.22.3-linux-x64.tar.xz cd /tmp && curl -fsSL https://nodejs.org/dist/v22.22.3/SHASUMS256.txt | grep ' node-v22.22.3-linux-x64.tar.xz$' | 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 activation guidance gives examples of when to trigger but does not define clear boundaries for when not to trigger. This ambiguity increases the chance of unnecessary tool invocation, which in this skill can cascade into package installation, network access, and token-configuration actions.