The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions to execute a bash script located at ${CLAUDE_PLUGIN_ROOT}/scripts/say. This script is called with a text argument to produce audio output using standard system tools like afplay or aplay.
[INDIRECT_PROMPT_INJECTION]: The skill processes a summary of accomplishments to provide audio feedback. If this summary includes content from untrusted external data (such as file contents or API responses), it could potentially be used for command injection if shell metacharacters are present in the summary.
Ingestion points: The summary text provided to the say script via shell arguments in SKILL.md.
Boundary markers: No explicit sanitization or escaping instructions are provided for the summary string.
Capability inventory: Shell execution via bash to run the say script.
Sanitization: The skill does not define specific sanitization or validation steps for the input text prior to shell execution.

voice-update