uxui-principles
Warn
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill's workflow instructs the user to execute "npx skills add uxuiprinciples/agent-skills", which downloads and installs code from a community-maintained repository (github.com/uxuiprinciples/agent-skills).
- [COMMAND_EXECUTION]: The operational workflow requires running shell commands via "npx" to fetch and configure the necessary tools for UX evaluation, representing arbitrary command execution from external sources.
- [DATA_EXFILTRATION]: The skill provides an optional configuration to connect to the "uxuiprinciples.com" API for "enriched output," which involves sending user-supplied interface descriptions and related data to a remote external server.
- [PROMPT_INJECTION]: The skill processes untrusted user input, creating a surface for indirect prompt injection attacks.
- Ingestion points: Interface descriptions, screen layouts, and user-provided interaction flows are ingested for analysis as described in the SKILL.md documentation.
- Boundary markers: There are no explicit instructions or delimiters telling the agent to treat the input as inert data rather than active instructions, which could allow malicious commands in the UI description to be obeyed.
- Capability inventory: The execution environment has the capability to run shell commands (npx) and make external network requests to the uxuiprinciples.com domain.
- Sanitization: The instructions do not define any validation, escaping, or sanitization steps for the user-provided content before it is processed or transmitted.
Audit Metadata