hunt-dispatch

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (critical risk: 1.00). The document explicitly orchestrates offensive cyber operations—fingerprinting targets, mapping to platform-specific "attack" skills, and instructing credentialed authenticated tests—constituting clearly malicious attack guidance.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill executes a runtime curl against "https://$TARGET" which fetches remote headers used to fingerprint the target and determine which platform skills to load, thus directly influencing agent behavior.