test(eeg): add deterministic synthetic Raw fixture (5 ch, 256 Hz, 10 s)

docs(bbb): fix stale CSV mentions in run_pipeline + module docstrings

docs: add README with quick start, status, and Day-2 onboarding map

feat(bbb): switch processed output to Parquet for dtype preservation

refactor(bbb): pin LF lineterminator; move DEFAULTs to top; reject dir output

feat(bbb): add run_pipeline orchestrator + CLI entrypoint with idempotent writes

fix(bbb): guard empty input; truncate index log; add KeyError + log tests

feat(bbb): expand SMILES → Morgan FP into model-ready DataFrame with drift logging

refactor(bbb): migrate to MorganGenerator API; consolidate test imports

feat(bbb): add Morgan fingerprint extraction with shape/dtype guarantees

refactor(bbb): tighten is_valid_smiles signature; document RDLogger global effect

chore: ignore all .venv*/ and venv*/ directories

feat(bbb): add SMILES validity guard with RDKit + test fixture

fix(core): apply level + propagate every call; tolerate pre-attached handlers

test(core): capture handler stream directly instead of monkey-patching caplog

feat(core): add shared structured logger with idempotent handler attach

chore: lock Python to 3.10–3.12 to match cp312-capped wheel availability

chore: pin runtime + dev dependencies for all three modalities

docs: tighten AGENTS.md — clarify Data Readiness checklist and storage boundary

docs: add AGENTS.md with vision, layout, standards, data readiness rules

chore: track planning docs and ignore .sixth/ tooling dir

chore: bootstrap NeuroBridge project skeleton