NILESH1003/codeguardian-env

Format episode score with 3 decimals for strict validation.

Fix strict reward and score clamping

fix: clamp reward scores to strictly (0, 1)

Update environment.py

Fix phase 2 task score bounds via np.clip

Fix score boundaries in inference baseline to strictly enforce range (0, 1)

fix: clamp grader score strictly between 0.01 and 0.99 for phase 2 criteria

Add uv lockfile and server entrypoints

Add pyproject.toml for OpenEnv multi-mode validation

trigger rebuild

fix: add HF yaml config

Initial commit: CodeGuardian OpenEnv submission