fix: fix SYSTEM_PROMPT KeyError crash in train_grpo.py

docs: update README and Blog with real Qwen3 benchmark results

fix: add all asset plots and metrics to repo (remove gitignore exclusions)

feat: add benchmark.py (Qwen3 vs baseline) and multi-model eval script

docs: add Blog.md writeup for hackathon submission

README file changes

fix: update Colab notebook to clone from HF Space repo

fix: add HF Spaces configuration header to README

feat: LogSentinel v2 — Multi-Agent SOC War-Room

Clamp task scores strictly within (0, 1)

Make inference.py bulletproof against network failures

Fix Dockerfile: use server.app:app instead of server:app

Fix multi-mode deployment validation

Add pyproject.toml for multi-mode deployment validation

Fix env variable names to match submission checklist

Fix inference.py output format to match spec exactly

Add root GET / endpoint for HF Spaces homepage

Deploy LogSentinel environment to HF Spaces

initial commit