Engram Explorer icon

Engram Explorer

Explore, create, update, and query the Engram knowledge graph

Actions25

Entity → Search

AI-generated

Summary

Searches entities in the selected Engram storage backend using a required group ID and query. Optionally combines text and semantic retrieval, filters by relevance and creation date, and returns either matching entities or an aggregate retrieval diagnostics record.

Inputs

  • Backend — Storage backend to search: Embedded (Graphology, default) or Neo4j (Remote).
  • Custom Storage Path — For the embedded backend, optional directory for graph data files; empty uses n8n’s default storage directory.
  • Group ID (required) — Session or group identifier that scopes the entity search.
  • Query (required) — Text used to find matching entities.
  • Search Mode — Text Only (default) uses full-text matching. Hybrid combines full-text and embedding-based semantic search with reciprocal-rank fusion.
  • Embedding Model — Model for semantic search when Hybrid mode is selected; uses the Engram Extraction LLM credential.
  • Limit — Maximum number of results; defaults to 10.
  • Min Relevance Score — Minimum relevance score from 0 to 1; defaults to 0.
  • Retrieval Diagnostics — When enabled, returns one aggregate record with bounded candidate decisions and a context-budget preview; disabled by default.
  • Diagnostics Candidate Limit — Maximum candidate decisions in the audit trace when diagnostics are enabled; defaults to 100.
  • Context Preview Token Budget — Token budget for the context preview when diagnostics are enabled; defaults to 1000.
  • Created After — Optional lower bound for entity creation date.
  • Created Before — Optional upper bound for entity creation date.

Output shape

A list of matching entity records, or one aggregate diagnostics record when retrieval diagnostics are enabled.

The result count is capped by Limit. Hybrid mode requires an embedding model. Exact record fields and diagnostics contents are not established by the supplied source view.

Examples

Example 1: Find entities in a group with text matching

Set Group ID and Query; leave Search Mode as Text Only.

Example 2: Find semantically related entities

Select Hybrid mode and an Embedding Model, then optionally tune the result limit and minimum relevance score.

Links

Discussion