terminal-diagrams
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a utility designed to help agents generate and validate terminal-based diagrams. It provides comprehensive instructions for handling Unicode width complexities and ANSI styling safely.
- [COMMAND_EXECUTION]: The skill utilizes a bundled script,
scripts/validate-layout.py, to verify diagram formatting. Analysis of the script confirms it is a benign utility that uses standard Python libraries to calculate visible cell widths and validate ANSI SGR sequences. It does not perform network operations, access sensitive files, or attempt privilege escalation. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it ingests user-provided diagram content and layout requests. However, the associated validation script only performs structural analysis of text (width, character properties) and does not interpret the content as executable commands or instructions, rendering the surface safe.
- Ingestion points: User-supplied layout requests defined in
assets/layout-request.jsonand diagram text generated based on user prompts. - Boundary markers: None implemented for the input diagram text.
- Capability inventory: Local file read and text-processing via
scripts/validate-layout.py. - Sanitization: The validation script includes strict filtering of unstable Unicode sequences (e.g., ZWJ, regional indicators) and non-SGR ANSI escape codes to prevent terminal manipulation.
Audit Metadata