lov-install-design

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands using npm and npx to install design components and run build scripts. These operations are core to the skill's primary purpose of design system installation.
  • [EXTERNAL_DOWNLOADS]: The skill downloads vendor-specific assets and open-source UI libraries (e.g., shadcn) from standard registries like NPM. These downloads are performed as part of the intended installation workflow.
  • [DATA_EXFILTRATION]: The skill uses scripts/profile_store.py to manage persistent user and brand configuration in ~/.config/agent-skills/. The script incorporates proactive security measures, specifically a blocklist for sensitive keys such as 'token', 'secret', 'password', and 'api_key' to prevent accidental credential exposure.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project files and branding specs to generate code updates, creating a potential surface for instructions embedded in external data.
  • Ingestion points: Project source files and branding specifications (SKILL.md).
  • Boundary markers: The workflow explicitly separates analysis and preview from actual modification phases to ensure user oversight (SKILL.md).
  • Capability inventory: File read/write access and shell execution for package management and builds.
  • Sanitization: Profile management includes keyword filtering and requires explicit confirmation for persistence operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 09:47 AM
Security Audit — agent-trust-hub — lov-install-design