The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references an external installation guide hosted on GitHub for the inference.sh CLI (belt).
Evidence: Link to https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md found in the Quick Start section of SKILL.md.
[COMMAND_EXECUTION]: The skill utilizes the belt CLI tool via the Bash tool to perform remote application execution and voice transformation.
Evidence: Multiple examples in SKILL.md use belt app run elevenlabs/voice-changer to process audio data.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of audio URLs and format strings which are passed as arguments to the CLI tool.
Ingestion points: The audio, voice, and model parameters within the --input JSON object in SKILL.md.
Boundary markers: The skill uses JSON structured input to encapsulate parameters, which provides a basic boundary but does not validate the content of the remote audio source.
Capability inventory: The skill uses belt app run to execute remote AI models and redirect output to local files (e.g., > professional.json in SKILL.md).
Sanitization: No explicit validation or sanitization of the provided URLs or input strings is performed before they are passed to the shell command.

elevenlabs-voice-changer