The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes a local shell script (adjust.sh) using node:child_process.execFile to handle video processing tasks. The use of execFile with arguments passed as an array effectively prevents shell injection vulnerabilities.
[INDIRECT_PROMPT_INJECTION]: The skill processes external media files, which represents a potential attack surface for indirect prompt injection. 1. Ingestion points: The input parameter in index.js accepts user-provided file paths. 2. Boundary markers: No explicit boundary markers are present to delimit untrusted data. 3. Capability inventory: The skill can execute local shell commands and perform filesystem read/write operations. 4. Sanitization: The speed rate is strictly clamped to a safe range (1.0 to 1.2), and the FFmpeg command utilizes the file: protocol prefix to prevent server-side request forgery (SSRF) and protocol-switching attacks.

speed