icon-set-extend
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run two local Python scripts,
scripts/infer_spec.pyandscripts/validate_icons.py. These scripts only use Python standard libraries and perform geometric analysis of SVG files. - [DATA_EXPOSURE]: The skill reads
.svgfiles from a user-provided directory to infer styling rules. It does not access sensitive system paths or credentials. - [REMOTE_CODE_EXECUTION]: All code is local and provided within the skill's scripts directory. There are no patterns of downloading or executing remote code.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted SVG data. While this is an ingestion surface, the processing is limited to XML parsing and geometric sampling for design purposes, which carries negligible risk for prompt injection under the provided scripts.
Audit Metadata