End-to-end pipeline deployment on EEE infrastructure

[arishofmann]

The pipeline already generates BenchmarkCards end-to-end. The next step is getting it deployed on EEE infrastructure so it can run automatically when new benchmarks come in.

Steps:

• Get access to EvalEval HF org and GitHub org
• Test HF inference credits, check which models are available
• Create HF dataset for generated cards: https://huggingface.co/datasets/evaleval/auto-benchmarkcards
• Validate that card quality holds with the new LLM backend
• Look at the existing EEE webhook/bot infrastructure and figure out how to plug in
• Set up automation so new benchmarks without a card get one generated and submitted as a PR
• Surface the confidence flags from validation in the PR so reviewers know which fields need the most attention