The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using the videoclaw CLI to initialize directories and set configuration values (videoclaw init and videoclaw config). These actions are performed to set up the project environment as requested by the user.
[INDIRECT_PROMPT_INJECTION]: The skill captures a project name from the user and interpolates it into a shell command (videoclaw init <project-name>). This represents a potential command injection surface if the input is not sanitized by the underlying agent platform.
Ingestion points: Project name collected via AskUserQuestion in SKILL.md.
Boundary markers: No explicit delimiters or instructions are provided to the agent to sanitize or escape the project name.
Capability inventory: Shell command execution via the videoclaw utility as seen in SKILL.md.
Sanitization: No specific sanitization logic is defined within the skill instructions.

video-init