ui-design-audit
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists exclusively of markdown instructions and checklists. It contains no executable scripts, binaries, or automated code functions.
- [PROMPT_INJECTION]: The skill includes no instructions that attempt to override agent behavior or bypass safety filters. It possesses an indirect prompt injection surface:
- Ingestion points: UI components in
src/components/,packages/ui/and configuration files liketailwind.config. - Boundary markers: None specified.
- Capability inventory: Limited to writing markdown files to the
plans/directory. - Sanitization: None specified for ingested codebase content. This surface is evaluated as safe due to the restricted output format and lack of dangerous capabilities like network access or shell execution.
- [DATA_EXFILTRATION]: No network access or data transmission is performed. The skill only reads project-specific UI code and configuration files. It does not target sensitive user data or credentials.
- [COMMAND_EXECUTION]: The workflow relies on file system reads and writes but does not execute shell commands, subprocesses, or dynamic code.
- [NO_CODE]: This skill provides instructions for the agent to follow and does not include any accompanying scripts or logic files.
Audit Metadata