The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructions include a command to install the CLI using curl -fsSL https://cli.inference.sh | sh. This pattern executes a remote script directly in the shell without prior verification, which is a high-risk operation even when sourced from the vendor's domain.
[EXTERNAL_DOWNLOADS]: The installation script downloads binary files from dist.inference.sh. This introduces external, unverified executable code into the agent's environment.
[COMMAND_EXECUTION]: The skill uses the Bash tool to run infsh commands. This allows the agent to execute subprocesses and interact with system resources to perform its functions.
[DATA_EXFILTRATION]: User-provided prompts and lyrics are sent to the inference.sh API for processing. Any sensitive data included in these inputs will be transmitted to the vendor's servers.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it interpolates user-controlled music descriptions into CLI commands.
Ingestion points: The --input JSON payload containing prompt and lyrics fields in the infsh command in SKILL.md.
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat user input as non-executable data.
Capability inventory: The skill allows shell command execution and remote API interaction via the infsh tool.
Sanitization: The skill lacks visible sanitization or validation mechanisms for the user-provided strings before they are passed to the shell.

ai-music-generation