better-icons

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes the better-icons CLI tool for icon searching, retrieval, and environment configuration.
  • [DATA_EXFILTRATION]: Includes the scan_project_icons tool which reads contents of project files to identify used icons, potentially exposing local code to the agent.
  • [EXTERNAL_DOWNLOADS]: Connects to https://api.iconify.design to retrieve icon assets and metadata. This is a well-known service for icon management.
  • [PROMPT_INJECTION]: Susceptible to indirect prompt injection due to untrusted data processing (Category 8).
  • Ingestion points: Project files are read via scan_project_icons and external data is fetched via search_icons from an API.
  • Boundary markers: No explicit instructions to ignore or delimit instructions embedded in the ingested data are present.
  • Capability inventory: File system writes to absolute paths via sync_icon, subprocess execution of the better-icons CLI, and network communication with the Iconify API.
  • Sanitization: No evidence of input validation or content sanitization is provided for file paths or retrieved icon content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 11:32 PM
Security Audit — agent-trust-hub — better-icons