design-import

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). Source: the user supplies a Claude Design handoff URL or local file via /ork:design-import, which is ingested at runtime by the ork:claude-design-orchestrator subagent that fetches the bundle and reads README.md and chats/*.md from the tarball.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill accepts HTTP(S) bundle URLs (e.g., https://claude.ai/design/abc123) and at runtime the orchestrator agent will WebFetch, extract, and parse the fetched tarball whose contents directly drive the agent's parsing/scaffolding decisions.