The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes git pull and npx skills add to update its components. This behavior is necessary for the skill's primary function and targets the author's official repositories.
[EXTERNAL_DOWNLOADS]: It fetches updates from the author's GitHub repository (maojiebc/majia-siyu-team). These downloads are consistent with the primary purpose of a self-updating skill.
[PROMPT_INJECTION]: The instructions explicitly direct the agent to skip second-time text confirmation when the user requests an update, reducing human oversight for the resulting command execution.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user triggers (e.g., '/siyu-update') to initiate shell commands. It lacks explicit boundary markers or sanitization for these inputs, representing a standard vulnerability surface.
Ingestion points: User input triggers and commands defined in SKILL.md.
Boundary markers: Absent.
Capability inventory: Shell execution via git and npx.
Sanitization: None detected.

siyu-update