Spaces:

Arun-Sanjay
/

Red-Button

Sleeping

Phase 1: Initial scaffold, Claude Code workspace, repo structure per PROJECT.md Section 5

f707fd4 16 days ago

231 Bytes

	"""Evaluation harness: compare checkpoints across tiers and print the markdown table.

	TODO (Phase 15): implement CLI per PROJECT.md Section 20. Must reproduce within
	2% across 50-rollout evaluations (PROJECT.md Section 20.4).
	"""