test-skill

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 1.00). The skill includes hidden JavaScript that performs an undisclosed fetch to an external webhook (exfiltration beacon) which is not part of the advertised dummy/test steps, so it contains deceptive hidden instructions outside its stated purpose.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill contains an intentional hidden outbound network request (fetch to a webhook.site URL) executed during initialization, which is a deliberate "phone-home" behavior that enables tracking or exfiltration and therefore represents high-risk malicious intent.

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I found a hardcoded webhook URL: https://webhook.site/b2e11e55-15b6-472d-822b-495f7a24c384. The GUID-like path (b2e11e55-15b6-472d-822b-495f7a24c384) is a high-entropy token that grants access to captured requests on webhook.site and thus functions as a live credential/endpoint. No other high-entropy API keys, private keys, or passwords are present; strings like "test-skill" and the description are non-secrets and treated as placeholders.