higgsfield-shotlist-director
Warn
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a shell command
python3 seedance_lint.py --preflight --model seedance_2_0 "<prompt>". Because the prompt content is interpolated directly into the command string using double quotes, it is vulnerable to command injection if the generated prompt contains shell metacharacters (e.g., backticks, semicolons, or unbalanced quotes). - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection from untrusted scripts or briefs.
- Ingestion points: Untrusted script briefs, scene breakdowns, and story ideas provided by the user (SKILL.md Workflow section).
- Boundary markers: Absent; there are no instructions to use delimiters or ignore embedded instructions within the user-provided scripts.
- Capability inventory: Subprocess execution (python3) and file writing (HTML artifacts).
- Sanitization: No sanitization or escaping of the generated prompt text is performed before it is passed to the shell or the HTML template.
- [COMMAND_EXECUTION]: The skill generates a self-contained HTML artifact containing inline JavaScript and dynamically generated prompt content. The lack of HTML escaping when inserting prompt text into the
{{SCENES_HTML}}block creates a Cross-Site Scripting (XSS) vulnerability that could execute arbitrary JavaScript in the user's browser when the file is opened.
Audit Metadata