The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions to install the Higgsfield CLI using a piped remote execution pattern: curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh. This method executes a script directly from a remote URL, which is a significant security risk as the script content could be modified at the source.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute higgsfield product-photoshoot create. User input collected during the pre-generation interview is interpolated directly into the shell command's arguments: --prompt "<short user-intent description from interview answers>". This creates a risk of command injection if the user provides input containing shell metacharacters like backticks or semicolons.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its ingestion of untrusted user data.
Ingestion points: User-provided descriptions and answers to interview questions in SKILL.md are used as CLI arguments.
Boundary markers: The instructions wrap the interpolated user input in double quotes ("$USER_INPUT"), which provides minimal protection against escaping shell arguments.
Capability inventory: The skill has access to the Bash tool to execute system commands.
Sanitization: There is no evidence of sanitization, escaping, or validation of the user-provided strings before they are passed to the shell.

higgsfield-product-photoshoot