fix: clamp ALL score outputs to (0.01, 0.99) — inference.py score + environment total_reward

fix(grader): clamp scores to (0.01, 0.99) — platform requires strictly between 0 and 1

fix: replace medium task with IDOR, tighten hard task grading

feat: multi-step env, pickle deserialization hard task, rebalanced difficulty

chore: production standardisation of the server codebase

Final submission readiness: cleanups, checklist, strict grader fix

chore: final calibration and validator fixes for OpenEnv submission

refactor: calibrate scores for progression Easy > Medium > Hard

feat: finalize OpenEnv alignment and calibrate rewards for QA

Add Java rate limiter and C++ event dispatcher tasks; update requirements

feat: add interactive web dashboard for space

refactor: update relative imports to resolve pathing issues after project restructuring

chore: initialize python virtual environment and dependencies