README: document self-contained task images (no external base image/repo/token)

README: fully agent-neutral (no agent names/specifics), generic -a <agent> run

README: drop AutoDS-specific section; add docker-less (-e singularity) run option for any agent

Reframe README: general benchmark for any Harbor agent; AutoDS specifics optional

README: present Docker (harbor) and venv as two explicit run options

Fix stale README/registry: single dataset, agent-neutral base instructions, data shipped on HF

Base instructions from original MLAgentBench repo; fix stale metadata (real data, 9 tasks)

full dataset incl. real data