The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides a script scripts/gen_vo.py that is intended to be executed to process voiceover generation tasks via the ElevenLabs API.
[CREDENTIALS_UNSAFE]: The utility script scripts/media_proxy.py reads a local credentials file at ~/.gooseworks/credentials.json. This is the standard method for the skill to authenticate with the vendor's own infrastructure and proxy services.
[DATA_EXFILTRATION]: The skill transmits diagnostic events, including the prompt text and performance metrics, to the vendor's logging endpoint at /api/internal/cli-logs to support monitoring and debugging of paid API calls.
[EXTERNAL_DOWNLOADS]: The skill downloads generated media assets (MP3 files) from the proxy service and potentially other media from the FAL media CDN.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests untrusted text input via the command line for voiceover generation.
Ingestion points: The --text argument in scripts/gen_vo.py processes external strings.
Boundary markers: None identified for the input text payload.
Capability inventory: The skill uses requests.post to interact with external APIs.
Sanitization: No sanitization or filtering is performed on the input text before transmission.

create-vo-elevenlabs