skills/kirklin/skills/unocss/Gen Agent Trust Hub

unocss

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOWEXTERNAL_DOWNLOADSDATA_EXFILTRATIONREMOTE_CODE_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION] (LOW): The documentation for content extraction (references/core-extracting.md) includes a configuration example using fetch() to retrieve remote text. If an agent is induced to incorporate malicious URLs into a project config, this could be used for data exfiltration during the build or extraction process.
  • [EXTERNAL_DOWNLOADS] (LOW): The skill describes the use of external CDNs (e.g., esm.sh) and the installation of numerous npm packages for presets and icon collections.
  • [REMOTE_CODE_EXECUTION] (LOW): The Icons preset (references/preset-icons.md) documents an autoInstall feature that automatically downloads and installs packages from npm. This capability could be exploited in a development environment if an attacker can control the icon collection names within the configuration.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 01:19 PM