obliteratus-abliteration

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The skill documents a technical toolkit for model abliteration, which involves removing safety guardrails and refusal mechanisms from open-weights language models.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs cloning an unofficial GitHub repository (https://github.com/elder-plinius/OBLITERATUS) and installing a third-party Python package (obliteratus), constituting an inbound operational dependency with weak/unknown provenance.