storyboard-generator
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
scripts/generate_image.py) to interface with the Gemini API for image generation. This script uses the standardurlliblibrary for network requests. - [DATA_EXFILTRATION]: The skill uses a hardcoded absolute file path (
/Users/ugreen/Documents/obsidian/) for output operations. Accessing specific absolute paths in a user's directory exposes local environment structure and presents a risk of unauthorized file writes if the path components (e.g., story names) are not sanitized. - [PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection by processing untrusted user stories and interpolating them into complex prompt templates for an image generation API.
- Ingestion points: User-provided story descriptions in Phase 1.
- Boundary markers: None; user content is directly concatenated into prompt strings.
- Capability inventory: File system write access and network access to Google APIs via the image generation script.
- Sanitization: No validation or escaping of user input is performed before usage in CLI arguments or API prompts.
Audit Metadata