higgsfield-product-photoshoot

Fail

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill includes instructions to download and execute a shell script from a remote URL (https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh) using the curl | sh pattern. This executes external code on the user's system without prior verification.
  • [COMMAND_EXECUTION]: The skill executes the higgsfield CLI tool by interpolating user-provided text into shell command arguments, specifically for the --prompt flag. This direct interpolation into a shell command environment creates a risk of command injection.
  • [EXTERNAL_DOWNLOADS]: The skill's setup process requires downloading an installation script from an external GitHub organization (higgsfield-ai).
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through its handling of user data.
  • Ingestion points: User responses from the pre-generation interview (e.g., product descriptions, mood, and styles) are used as arguments for the --prompt flag in SKILL.md.
  • Boundary markers: Arguments are enclosed in double quotes.
  • Capability inventory: The skill utilizes the Bash tool to execute system commands and manage image uploads.
  • Sanitization: No validation or escaping of the user-provided prompt strings is implemented before they are executed in the shell environment.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 20, 2026, 11:36 PM
Security Audit — agent-trust-hub — higgsfield-product-photoshoot