llm-icon-finder
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to construct URLs for downloading static image assets (SVG, PNG, WEBP) from the official LobeHub GitHub repository (
raw.githubusercontent.com/lobehub/lobe-icons). These downloads are limited to non-executable media files required for the skill's primary function. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface where it ingests user-supplied icon names and incorporates them into
curlcommands. - Ingestion points: User query for icon names in
SKILL.md. - Boundary markers: None explicitly defined in the instructions.
- Capability inventory: The skill uses the Bash tool with
curlfor downloading files. - Sanitization: No specific sanitization logic is described, but the operation is scoped to static asset retrieval from a specific repository.
- [SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected. All external dependencies and resource URLs point to well-known open-source libraries and services.
Audit Metadata