Consolidate fragmented entity aliases for the workspace (CORE-GRAPH-ALIAS-RESOLVE-2 B2).
Merges each unambiguous single-token alias node (e.g. "Hudson") into its multi-token canonical ("Rock Hudson") over the COMPLETE workspace graph, abstaining on collisions ("Hudson" with both "Rock Hudson" and "Rochelle Hudson" present). Run after a bulk ingest, when the full ambiguity picture is present. Workspace-scoped (tenant isolation).
Parameters:
dry_run: If true, report the resolve/abstain plan without mutating the graph.disambiguate: Opt-in (CORE-GRAPH-ALIAS-DISAMBIG-1, default false). Additionally recover colliding surfaces by structural typed-neighbor overlap — merge only on a confident, clear winner, else keep abstaining (never mis-merge). Extractor-dependent; helps LLM-extracted graphs.
Returns:
resolved: alias surfaces merged into a canonical (unambiguous + disambiguated)abstained: single-token surfaces left untouched (>=2 canonical candidates)redirected_edges: external edges moved onto canonicalsambiguous: the abstained surfacesdisambiguated: collisions recovered by the opt-in disambiguation pass (0 unlessdisambiguate=true)dry_run: echoes the request flag- plus scope context (
workspace_id,user_id)
Query Parameters
Response Body
application/json
application/json
curl -X POST "https://example.com/memory/graph/resolve-aliases"null{
"detail": [
{
"loc": [
"string"
],
"msg": "string",
"type": "string",
"input": null,
"ctx": {}
}
]
}Reset Pipeline Endpoint
Reset a pipeline, clearing its state. Requires authentication and operates on user-scoped data.
Resolve Conflict
Resolve a conflict between assertions. Auto-resolution cascade (if enabled): 1. Wikipedia lookup - verify against Wikipedia 2. LLM reasoning - ask GPT to fact-check 3. Grounding check - check existing provenance 4. Recency heuristic - prefer recent info for temporal conflicts Falls back to manual strategy if auto-resolution fails.