The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through user-provided inputs.
Ingestion points: The prompt and prompt_lyrics parameters defined in SKILL.md are sourced from user input.
Boundary markers: User inputs are wrapped in double quotes in the shell command template (--prompt "<style>" --prompt_lyrics "<lyrics>").
Capability inventory: The skill has the ability to execute the meitu command and write files to the ./output/ directory.
Sanitization: The instructions do not specify any validation, filtering, or escaping of user input before it is interpolated into the shell command.
[COMMAND_EXECUTION]: The skill executes external CLI commands to perform its core tasks.
It invokes the meitu command-line tool for audio generation and tool updates.
[CREDENTIALS_UNSAFE]: The skill accesses sensitive local credential files for service authentication.
It is configured to read ~/.meitu/credentials.json to retrieve MEITU_OPENAPI_ACCESS_KEY and MEITU_OPENAPI_SECRET_KEY.

audio-song-generate