fix(report): add explicit JSON structure guidance in prompt

fix: apply lazy init pattern and pydantic-ai fixes to ReportAgent (#12)

feat(phase7): implement Hypothesis Agent for mechanistic reasoning

fix: add type ignore comment for Agent instantiation in JudgeHandler

feat(phase3): implement judge slice (LLM Judge, Prompts, Models)

feat(phase1): Foundation & Tooling complete

adds scaffolding