Skills
skills/ibelick/ui-skills/improve-ui/Gen Agent Trust Hub

improve-ui

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted documentation such as DESIGN.md to establish audit criteria. Ingestion points: DESIGN.md and repository guidance. Boundary markers: Absent. Capability inventory: File-write restricted to design-plans/. Sanitization: Absent. Malicious instructions in these files could influence generated plans.
  • [COMMAND_EXECUTION]: The plan template includes a placeholder for a 'check command'. Although the skill is read-only and does not execute code, it can suggest commands based on analyzed data for downstream use.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 01:39 PM
Security Audit — agent-trust-hub — improve-ui