v2.0.18
2.0.18 (2026-07-18)
Bug Fixes
-
indexing: separate operation and storage IDs (61a1a84)
Separate transient operation identity from stable document IDs so concurrent requests cannot overwrite status or progress state.
Operation Tracking:
- Assign each add and reindex request an opaque operation UUID
- Key status and progress notification state by operation ID
- Preserve document IDs as compatibility identifiers in responses and statuses
Crawl Storage:
- Derive queue and dataset names from a SHA-256 hash of the normalized URL
- Remove both current hashed and historical hostname-derived datasets on delete