The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process user-provided text for speech generation, which creates a vulnerability surface for indirect instructions to influence agent behavior.
Ingestion points: The text parameter in the TTSSelector().execute() and DoubaoTTS().execute() method calls within SKILL.md serves as the entry point for untrusted data.
Boundary markers: No explicit delimiters, boundary markers, or specific instructions to ignore embedded commands within the text are defined in the skill documentation.
Capability inventory: The skill utilizes network capabilities to communicate with the Volcengine/ByteDance OpenSpeech API and possesses file system write access to save generated audio and metadata JSON files.
Sanitization: The instructions do not describe any sanitization, validation, or filtering of the input text before it is transmitted to the external API.

doubao-tts