claim-refinement
$
npx mdskill add yogsoth-ai/de-anthropocentric-research-engine/claim-refinementRefines claims to survive counterexamples while maximizing explanatory power
- Solves the problem of claims being invalidated by counterexamples
- Uses logical analysis and lemma incorporation techniques
- Evaluates counterexamples to narrow scope or split claims
- Returns a revised claim with preserved explanatory power
SKILL.md
.github/skills/claim-refinementView on GitHub ↗
--- name: claim-refinement description: "Propose a refined claim that survives counterexamples while preserving maximum explanatory power (Lakatos lemma-incorporation)." execution: subagent prompt: ./prompt.md used-by: [adversarial-stress-testing] --- # Claim Refinement Subagent that refines claims to survive counterexamples using lemma incorporation, scope narrowing, or claim splitting.