Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed video subtitle-removal integration, but users should expect qhkit installation and video uploads to the provider service.
Install only if you are comfortable using the qhkit cloud service, providing its API token, and uploading the videos you ask it to process. Prefer a reviewed or pinned qhkit version in controlled environments, and avoid adding Node or PATH changes permanently unless you intend to keep using the tool.
The description says subtitle-removal requests 'must trigger', using broad keywords that can match many normal user requests. Overly mandatory activation increases the chance the skill runs in unintended contexts, causing unnecessary file handling, tool execution, or external service uploads of user videos.
The skill instructs fallback execution via npx @iqinghu/qhkit <命令> ... without pinning an exact package version. That allows the runtime to fetch whatever version is current at execution time, which weakens supply-chain integrity and can introduce unexpected behavior or a compromised release into the agent environment.
The duplicated finding identifies the same persistence behavior: unpacking software into a user-controlled directory and prepending it to PATH, with optional persistence across shells. In an agent skill, this is security-relevant because it encourages durable environment modification as part of normal task execution, which is more dangerous than a one-off command.
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:$PATH"
The duplicated finding identifies the same persistence behavior: unpacking software into a user-controlled directory and prepending it to PATH, with optional persistence across shells. In an agent skill, this is security-relevant because it encourages durable environment modification as part of normal task execution, which is more dangerous than a one-off command.
cd /tmp && curl -fsSLO https://nodejs.org/dist/v22.22.3/node-v22.22.3-linux-x64.tar.xz cd /tmp && curl -fsSL https://nodejs.org/dist/v22.22.3/SHASUMS256.txt | grep ' node-v22.22.3-linux-x64.tar.xz$' | 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 trigger examples are short and permissive, but they do not define boundaries, disambiguation rules, or cases where the skill should not run. In a media-processing skill that may upload local files or contact remote APIs, ambiguous invocation logic raises the risk of accidental execution and unintended data transfer.