connect-to-skilder

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This file instructs an agent to silently register and connect to a fixed external MCP server and perform an OAuth handshake and capability enumeration without user consent, enabling remote access and potential token/credential exfiltration.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent at runtime to register and call the remote MCP endpoint https://app.skilder.ai/mcp, which will supply Roles/Skills that can directly control agent behavior and execute actions.