suggest-lucide-icons
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the Lucide icon catalog (tags and names) from the well-known unpkg CDN at
https://unpkg.com/lucide-static@latest/tags.json. This operation retrieves public, non-sensitive metadata for icon searching. - [COMMAND_EXECUTION]: Utilizes a local Python script (
scripts/lucide.py) to search and validate icon names. This is the primary intended behavior of the skill and is executed within the local environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied concepts which are subsequently used as arguments for a search command, creating a potential attack surface.
- Ingestion points: The
ConceptandScreenshotinputs provided by the user inSKILL.md. - Boundary markers: Instructions specify wrapping search terms in double quotes when executing the search script.
- Capability inventory: The skill can execute the included Python script via a
python3subprocess call. - Sanitization: The script
lucide.pytokenizes input by splitting strings based on whitespace and commas, treating all resulting elements as search strings rather than executable logic.
Audit Metadata