The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several shell commands (curl, jq, mkdir, echo, cat, rm) to manage local configuration in .listenhub/tts/config.json, construct API request bodies, and handle the downloading of generated audio files.
[EXTERNAL_DOWNLOADS]: Communicates with https://api.marswave.ai to perform text-to-speech synthesis and fetch speaker lists. These downloads are directed to the vendor's official API infrastructure and are necessary for the skill's functionality.
[DATA_EXFILTRATION]: Transmits user-provided text and script content to the external Marswave AI API. This data transmission is the core function of the skill and is transparently documented in the instructions.
[PROMPT_INJECTION]: The skill processes untrusted user text and script segments which are subsequently interpolated into JSON payloads and shell commands. While this presents an indirect prompt injection surface, it is a standard requirement for TTS services.
Ingestion points: User-supplied text and multi-speaker scripts in SKILL.md (Quick Mode and Script Mode).
Boundary markers: No specific delimiters or "ignore" instructions are defined for the input text when constructing shell commands.
Capability inventory: File system access (read/write config), network access (curl), and shell execution (bash, jq).
Sanitization: The instructions do not specify explicit sanitization or escaping mechanisms for the user input before it is passed to the API endpoints.

tts