fix: baseline always runs full rule sequence first, fix fill_nulls string mode

fix: fill_nulls mode on string columns (category/region), fix format crash

fix: GET /reset and /step return 200, POST /reset handles empty body

fix: real graders, real scores, /reset accepts empty body

Implement task scores and update scoring logic

Update graders.py

Update environment.py

Change grading functions to return fixed score

Update environment.py

Update graders.py

Enhance score clamping for all task branches

Fix score initialization and clamping in grading functions

Change initial previous score from 0.0 to 0.001

Update graders.py

Update default score values in environment.py

fix: all edge cases return 0.001 not 0.0, scores strictly (0,1)

fix: clip grader scores to strictly (0.001, 0.999) per validator requirement

fix: add openenv-core dep, project.scripts, main() entrypoint

fix: /reset accepts empty body from validator

error fix

DataClean OpenEnv v2.0 - Data Drift + Parallel Baseline