ops-voice

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly runs commands that print secret values (e.g., printenv, security -w, grepping shell profiles and Doppler outputs) and instructs to "Present all findings" and use those keys in curl headers, which requires the LLM/agent to handle and potentially output secrets verbatim.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill contains deliberate credential-harvesting behavior (auto-scanning environment variables, shell profiles, Doppler across all projects, Dashlane and macOS Keychain) and prints/validates secrets, which is high-risk data-exfiltration/credential-theft behavior; no obfuscated backdoor or remote shell was found, but the setup flow is explicitly designed to collect sensitive credentials and could be abused to exfiltrate them.