What problem does it solve?
This Skill helps you identify the single most damaging reason a hostile reviewer might give to reject your paper, then checks whether your current text truly neutralizes that critique.
Core Features & Use Cases
- Adversarial attack memo generation: Produces an approximately 200-word rejection paragraph that commits to one strongest failure mode (e.g., theorem validity, scope overclaim, assumption-vs-claim mismatch).
- Fresh point-by-point adjudication: Decomposes the attack into 3–7 atomic rejection points and classifies each as answered_by_current_text, partially_answered, or still_unresolved using evidence from the current paper files only.
- Actionable fixes discovery: Produces prioritized top action items (detect-only) to guide rebuttal preparation or the next improvement-loop round.
Quick Start
Use the kill-argument skill on your paper directory (for example, run it with your path in the paper-directory argument) to generate KILL_ARGUMENT.md and KILL_ARGUMENT.json in that same directory.