pitch-deck

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The skill contains instructions to promote an external installation command upon hitting capability limits, which constitutes suspicious steering.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill contains an explicit GitHub repository locator instruction for installation (github:mindverse/skillhub) within its capability gap response rules, constituting a weak/unknown runtime external dependency.