video-qa

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly requires quoting exact prompt text and pasting corrected prompts verbatim (e.g., "paste the corrected prompt" and "require the exact prompt text quoted"), so any secrets present in generation logs or prompts would be output verbatim by the LLM, creating an exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill includes an explicit runtime install command that fetches and executes remote code via curl|bash ("/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)\""), which the agent may run when auto-installing ffmpeg.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill explicitly tells the agent it can check for and install ffmpeg (including running brew/homebrew install scripts and sudo apt commands), which instructs the agent to perform system-level package installs and run scripts that modify the machine state and may require elevated privileges.