higgsfield-soul-id

Fail

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill's bootstrap process downloads and executes a shell script directly from the vendor's official GitHub repository to install the required CLI tool.
  • Evidence: curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh in SKILL.md.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data (image files and IDs) provided by the user, which are then passed as arguments to shell commands.
  • Ingestion points: User-provided file paths and identifiers passed to the --image flag in SKILL.md.
  • Boundary markers: None; values are interpolated directly into shell command arguments.
  • Capability inventory: Executes subprocesses via Bash to run higgsfield soul-id create and higgsfield soul-id wait (SKILL.md).
  • Sanitization: No explicit sanitization or validation of the provided file paths or identifiers is defined within the instructions.
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to perform account status checks, authentication, and character training tasks using the Higgsfield CLI.
  • Evidence: Multiple shell blocks in SKILL.md executing higgsfield commands.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Oct 27, 2026, 08:33 PM