Kill Argument Adversarial Review
Stress-tests a paper with a strongest-possible rejection memo, then a point-by-point defense.
- Stars
- 18,482
- Licence
- MIT
- Files
- 1 file
What it does
Once installed, the agent runs a two-stage adversarial review of a paper folder. A first fresh reviewer writes a single roughly 200-word rejection argument aimed at the paper's headline claims, the kind of sentence a senior area chair could use to reject it. A second fresh reviewer then defends the paper against each point, sorting each objection as answered, partially answered, or still unresolved. Results are saved as KILL_ARGUMENT.md and KILL_ARGUMENT.json, and an HTML version may be generated.
Great for
- Researchers preparing a theory paper for submission after it has already passed standard review rounds
- Authors anticipating reviewer-2 objections during rebuttal preparation
- Anyone who wants to find framing or scope overclaims that score-based reviews missed
Skip it if
- Empirical papers without theorems, where the skill's own guidance points to a different review skill (/research-review) (wrong tool for the job)
- Use inside a scheduled loop: the skill is designed to run once, and repeated runs add little signal unless the paper changed (wasted effort)
- Treating the verdict as ground truth: it is generated by language models and can be wrong or overly harsh (over-reliance risk)
Try asking
- “Run kill argument on ./my-theory-paper before I submit to the conference”
- “Simulate reviewer-2 for the headline claims in this paper and prepare rebuttal points”
- “Do an adversarial review of the paper directory to find unresolved critical issues”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
kill-argument.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip kill-argument.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip kill-argument.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Run it once after the draft is stable, not repeatedly on a timer, so each run reflects a real change in the paper.
- Give it a paper directory with the theorem statements and abstract, since the attack targets headline claims.
- Check each still-unresolved point yourself against the paper before acting on it, because the reviewers are models that may misjudge.
What's inside
- SKILL.md
Redistributed unmodified from wanshuiyin/Auto-claude-code-research-in-sleep under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.