QueryCells
Execute aCellQuery against the snapshot. Planner picks cheapest index; remaining predicates applied in-memory.
CellQuery fields
| Field | Description |
|---|---|
Query | Full-text search query |
RequireTags | Tags that must all be present (AND) |
AnyTags | Tags where at least one must be present (OR) |
ExcludeTags | Tags that must not be present (NOT) |
SourceID | Filter by source ID |
MinConfidence | Minimum confidence threshold |
MaxConfidence | Maximum confidence threshold |
After | Filter by ValidFrom (after this time) |
Before | Filter by ValidFrom (before this time) |
Center | Spatial query center coordinate |
Radius | Spatial query radius |
Embedding | Vector for ANN-accelerated seed selection |
MaxResults | Maximum results to return |
MaxScanRows | Maximum rows to scan |
SortBy | Sort order |
Explain | Return explanation string |
Sort order
| Sort order | Description |
|---|---|
SortByScore | Sort by composite score (default) |
SortByConfidence | Sort by confidence |
SortByRecency | Sort by creation time |
SortByCoord | Sort by coordinate |
Query planner index selection
The planner picks the cheapest index based on the query:| Condition | Primary index |
|---|---|
RequireTags non-empty | tag/ secondary index |
SourceID set | source/ secondary index |
After or Before set | time/ week-bucket index |
Center+Radius set | Ring walk around Center |
Embedding set | HNSW graph (ANN search) |
| Fallback | Full scan (ring walk from origin, radius 32) |
Temporal queries
After/Before filter on cell ValidFrom. Cells with no ValidFrom are excluded from any temporal query. Uses the existing time/ weekly-bucket index.