Update README with testing instructions and new config

Fix final phase 1 and phase 2 inference and scoring issues

readme

readme

fix: change body to empty dict to prevent Pydantic 422 Unprocessable Entity errors during validation

chore: update Dockerfile and inference.py based on local edits

chore: add test-all.sh bash script to validate everything at once

fix: update base Docker image to bypass registry errors and sync requirements path

fix: add root requirements.txt for Phase 2 validation and handle API keys/LLM crashes in inference.py

chore: push verified deployment fixes

docs: update README for expert tasks, dynamic max_steps, and reward tweaks

improve: hard tasks max_steps=10, fix reward gap, clearer descriptions

Add Expert tier tasks to evaluate multi-step UUID stateful tracking and OpenAPI discovery

Implement strict judge feedback: ambiguous task descriptions, server-side payload clues, increased exploration steps

Enhance episode state isolation and realistic grader hints for max scoring

Update inference.py environment variable parsing to match submission checklist

Fix submission validation configuration and dependencies

fix: add main(), project.scripts, uv.lock for openenv validate

Add MIT license

Final README with full documentation

fix: correct inference.py log format [START]/[STEP]/[END]

fix: remove Optional types to pass openenv validate

fix: replace Optional[dict] with Any for openenv validate

fix: Dockerfile to root, port 7860, README frontmatter

readme

Readme

Initial submission: API Debug Environment