office-web-ui-system
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill includes a local utility script,
scripts/scan_ui_locators.py, which helps the agent identify UI components in the user's project. The script performs local file reads and regex searches within the project directory using standard Python libraries, with no evidence of network communication or unauthorized file access.\n- [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface as it reads project source files (e.g., .vue, .tsx, .html) via the scanner script to identify UI components. The risk is categorized as safe because the script only extracts specific CSS tokens rather than processing arbitrary content, and the agent uses this metadata solely for layout orientation. Ingestion points: Local project source files. Boundary markers: Not explicitly defined in instructions. Capability inventory: Read-only access to source code files, no network or write permissions. Sanitization: Tokenization via specific regex and keyword matching.
Audit Metadata