dpai-craft

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill interacts with external interface source code and project design rules as part of its primary workflow.
  • Ingestion points: The workflow in SKILL.md describes reading project interface rules, design tokens, and source code for components and screens.
  • Boundary markers: The skill does not provide specific instructions to delimit untrusted project content or ignore embedded instructions.
  • Capability inventory: The skill utilizes get_project_context and verify_preview tools and has the capability to implement source code changes.
  • Sanitization: There are no explicit instructions for sanitizing or validating content retrieved from the external project environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 02:02 AM
Security Audit — agent-trust-hub — dpai-craft