The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands via the Bash tool. Specifically, it uses node ${CLAUDE_PLUGIN_ROOT}/src/prompt-coach.js with user-supplied text interpolated directly as a command-line argument.
[PROMPT_INJECTION]: The skill contains a surface for indirect prompt injection and command injection. It takes untrusted user input and places it within a shell command string.
Ingestion points: The user's last natural-language prompt or text provided as an argument to the /cco-coach command.
Boundary markers: The instructions suggest wrapping the input in double quotes ("the prompt text here"), but do not provide guidance on escaping shell metacharacters or handling input that contains quotes.
Capability inventory: The skill is granted access to the Bash tool, allowing it to execute arbitrary shell commands on the host environment.
Sanitization: There are no instructions or logic provided to sanitize, validate, or escape the user-provided content before it is processed by the shell.

cco-coach