Structured ML hypothesis investigation for Claude Code — ensemble critique (default) or adversarial debate, empirical testing, peer review, and coherence audit. Benchmark: 3× independent critics outperform structured debate on regular cases; adversarial exchange earns its keep when the question is empirically ambiguous.
machine-learning hypothesis-testing peer-review ml-agents llm agentic-workflow claude-code-plugin empirical-validation research-workflow ensemble-critique
-
Updated
Apr 19, 2026 - Python