higgsfield-mcp

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute the shell command open "outputs/" on macOS systems to reveal generated files to the user.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its web scraping capabilities.
  • Ingestion points: The show_marketing_studio tool (documented in SKILL.md and prompting/prompt-library/marketing-studio.md) scrapes data from external product URLs provided by the user.
  • Boundary markers: There are no instructions for the agent to use delimiters or ignore embedded instructions within the scraped product data.
  • Capability inventory: The skill allows for media generation (generate_video, generate_image), character training (soul_train), and directory opening (open).
  • Sanitization: The skill does not specify any validation or sanitization steps for the data retrieved from external product pages before it is incorporated into video/image generation prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 06:33 PM
Security Audit — agent-trust-hub — higgsfield-mcp