Frontend UX
Warn
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill is configured to fetch its primary content from a remote URL (
https://ethskills.com/frontend-ux/SKILL.md) during the installation phase via thecurlcommand. - [REMOTE_CODE_EXECUTION]: The setup instructions in
catalog.jsondirect users to install a plugin from a non-trusted GitHub repository (https://github.com/austintgriffith/ethskills) and use a custom installer (clawhub). This introduces code from a third-party source into the agent's environment without automated verification. - [INDIRECT_PROMPT_INJECTION]: The skill uses an external ingestion point where it reads instructions from
ethskills.comand instructs the agent to follow these rules before building UI. - Ingestion points: Content retrieved from
https://ethskills.com/frontend-ux/SKILL.md. - Boundary markers: None present to distinguish external instructions from system constraints.
- Capability inventory: The installed plugin from GitHub possesses the full execution capabilities of the agent's plugin system.
- Sanitization: No validation or sanitization is performed on the remote markdown content before it is processed by the agent.
Audit Metadata