obliteratus
OBLITERATUS enables the removal of refusal behaviors from LLMs using advanced mechanistic interpretability techniques without retraining.
Install this skill
or
99/100
Security score
The obliteratus skill was audited on May 21, 2026 and we found 1 security issue across 1 threat category. Review the findings below before installing.
Categories Tested
Security Issues
low line 28
External URL reference
SourceSKILL.md
| 28 | https://www.youtube.com/watch?v=8fG9BrNTeHs ("OBLITERATUS: An AI Agent Removed Gemma 4's Safety Guardrails") |
Scanned on May 21, 2026
View Security Dashboard