Rigorous LLM research in two layers: a multi-agent dialectic synthesis prompt + an agent-loadable research skill. Hierarchy of evidence, verbatim citation gates, no fabrication.
research multi-agent evidence-synthesis llm prompt-engineering llm-research cursor-rules adversarial-ai confidence-scoring claude-skill epistemics epistemic-calibration hierarchy-of-evidence
-
Updated
May 24, 2026 - JavaScript