unified-ai-gateway

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly pulls and later registers/runs the container image ghcr.io/happy520ai/unified-ai-system/mcp-server@sha256:751a0d32acd2d6b1da6ad9ac67987fbd1ff36ce26b7160014d8605f18b7907b3 during setup/registration, which fetches remote code the agent depends on and can be executed by the runtime.