fix: update score range to 0.1-0.9 per hackathon spec, add inference tests, update validator

fix: remove early break in run_campaign - ensure all 3 tasks run for grading

fix: Add LLM safety timeout and graceful exit handling

feat: Add Phase 2 validation - multi-task graders with partial credit scoring

fix(deploy): Implement dummy server to satisfy HF running state & POST requirements

Update to Qwen2.5-Coder-32B-Instruct and add keep-alive loop for persistent Space

Update model to Qwen2.5-Coder-32B-Instruct

Update model to Qwen2.5-7B-Instruct and add explicit tool schemas to system prompt

Update default model to Qwen2.5-Coder-32B-Instruct

fix: switch to HF's new router.huggingface.co endpoint (deprecated api-inference)

fix: make HF_TOKEN mandatory at startup per hackathon requirements

fix: switch from deprecated HF API to Together AI API (free tier)

fix: improve HF_TOKEN error handling - print setup instructions once instead of every step

fix: make HF_TOKEN optional at startup, validate at runtime for Space compatibility

fix: restore HF_TOKEN as mandatory per hackathon requirements

feat: switch to Qwen2.5-3B-Instruct with HF Inference API + optimize for 2vCPU/8GB RAM

Initial commit: OpenEnv-CloudSOC benchmark environment