PlateerLab/synaptic-memory

tools

• agent_search

  FTS + vector hybrid search with intent routing

• aggregate_nodes

  GROUP BY + COUNT/SUM/AVG/MAX/MIN with optional WHERE pre-filter

• compare_search

  Auto-decompose multi-topic queries, search in parallel

• count

  Structural count by kind/category/year

• deep_search

  Recommended. Search → expand → read documents in ONE call

• expand

  1-hop graph neighbours

• filter_nodes

  Property filter (>=, <=, contains) — returns {total, showing} for accurate counting

• follow

  Walk a specific edge type

• get_document

  Full document with query-relevant chunks

• join_related

  FK-based related record lookup — walks RELATED edges (O(degree))

• knowledge_add_chunks

  BYO-chunker path for pre-split content

• knowledge_add_document

  Ingest a long-text document with automatic sentence-boundary chunking

• knowledge_add_table

  Ingest structured rows → ENTITY nodes + FK edges

• knowledge_backfill

  Repair graphs missing embeddings or phrase hubs (v0.14.4+)

• knowledge_ingest_path

  Ingest a CSV / JSONL / text file from the local filesystem

• knowledge_remove

  Delete a single node with edge cascade

• knowledge_search

  Core semantic search (routes through EvidenceSearch in v0.14.2+)

• knowledge_sync_from_database

  Incremental sync from a live database (CDC)

• list_categories

  Category list with document counts

• search_exact

  Literal substring match for IDs/codes

• session_info

  Multi-turn session state

• top_nodes

  Single-call top-N ranking — "가장 X한" / "top N" / "최대/최소" / "최근" questions without composing aggregate_nodes. Each row carries sort_value for chaining into join_related / filter_nodes(from_ids=...). v0.