antigravity-skill-orchestrator

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

• Third-party content exposure detected (medium risk: 0.30). The skill-orchestrator reads the user’s free-text request (Step 1: “Read the user's request” and “Use the user request as input”) to decide which skills to invoke, so an outsider can inject content directly into the workflow the LLM ingests.