[anandpant]

Summary

• split the repo into a real Cargo workspace with convex-cdc-core, convex-target-s3, and convex-sync
• move AWS and Databricks assets under platform/ with explicit S3 and Databricks-native target ownership
• add Databricks-native bootstrap, Lakeflow AUTO CDC template, and extractor job assets
• harden the Databricks S3-backed view sync with external-location preflight checks
• tighten the README and durable docs, remove stale planning docs, add a release-artifacts guide, and drop the repo-local AGENTS.md

Verification

• cargo fmt --check
• cargo clippy --workspace --all-targets --all-features -- -D warnings
• cargo test --workspace
• python -m py_compile platform/databricks/native/extractor/convex_cdc_job.py
• live convex-sync sync-once against the configured Convex deployment
• live convex-sync run --max-iterations 1 end-to-end through S3 publish
• live Databricks S3-backed view registration against the real prod prefix
• live Databricks-native bootstrap SQL apply / checkpoint table smoke test
• live Databricks serverless notebook smoke run for the native extractor

Notes

• release automation itself is intentionally left for a follow-up PR; this PR only defines the current release artifact shape in docs/release-artifacts.md
• the Databricks-native extractor was debugged through several real serverless notebook runs until it completed successfully

---