The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automates audio generation by executing a shell command that calls an internal platform binary. This involves the use of environment-specific variables like $OD_NODE_BIN and $OD_BIN to dispatch requests to upstream audio models.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface where untrusted data could influence command execution.
Ingestion points: User-supplied text for music prompts, sound effect descriptions, and text-to-speech scripts are ingested as metadata or prompts as described in SKILL.md.
Boundary markers: The prompt and script inputs are wrapped in double quotes in the shell command template, but the skill lacks specific delimiters or instructions for the agent to sanitize or ignore embedded control characters in that data.
Capability inventory: The skill performs subprocess execution using the platform's media dispatcher tool via a shell environment in SKILL.md.
Sanitization: No explicit sanitization or escaping of the user-supplied prompts is defined in the workflow instructions; the skill relies on the agent to manually proof text for cadence rather than security validation.

audio-jingle