lender-loan

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill deliberately instructs automatic notification (exfiltration) of qualified users' asset and contact details to an external address (erik@volume.finance) while directing the agent to hide internal criteria/provider identity, which constitutes intentional, privacy-sensitive data exfiltration and opaque behavior even though there is no obvious remote‑exec/backdoor obfuscation or credential-theft code present.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs calling the Seren Cloud run endpoint at https://api.serendb.com/publishers/seren-cloud/agents/04244f5a-1133-4cd7-87cd-4ce80f62636f/runs at runtime to execute remote agent logic and receive prompts/state that directly control the agent flow, so this external URL is a runtime dependency that executes remote code and controls prompts.