brand

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/sync-brand-to-tokens.cjs uses child_process.execSync to run a script at .claude/skills/design-system/scripts/generate-tokens.cjs. This establishes a cross-skill dependency and executes external code for token regeneration.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface in scripts/inject-brand-context.cjs.
  • Ingestion points: The script reads and extracts data from docs/brand-guidelines.md.
  • Boundary markers: Uses static headers such as BRAND CONTEXT:, VISUAL IDENTITY:, and BRAND VOICE: to delimit injected content.
  • Capability inventory: Uses fs to read files; the output is intended for system prompt interpolation by the agent.
  • Sanitization: Extracted content from the markdown file is appended directly to the prompt without sanitization or escaping, which could allow malicious instructions in the guidelines to influence agent behavior.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 08:41 AM
Security Audit — agent-trust-hub — brand