The Agent Skills Directory

[COMMAND_EXECUTION]: The skill processes user-provided input (<name>) and interpolates it directly into multiple Bash commands, including git branch --list <name>, ls <BASE_DIR>/<name>, and git worktree add <BASE_DIR>/<name> -b <name>.
Ingestion points: The <name> argument is provided by the user (as seen in SKILL.md).
Boundary markers: None are present to prevent shell metacharacters (like ;, |, or &) from breaking out of the intended command.
Capability inventory: The skill uses the Bash tool to execute git and filesystem commands, as defined in SKILL.md.
Sanitization: There is no evidence of input validation, escaping, or sanitization before the user input is executed in the shell context. If dangerouslyDisableSandbox is enabled as suggested in the text, the impact of a command injection could extend to the host filesystem.

mkwt