Tool reference · Knowledge graph & temporal facts
memory_entity_resolve
Maps a mention such as "Postgres" or "pg" to one canonical entity id within a project and type, creating the entity if it is new.
local stdio server writes idempotent
Caution. Embedding matches need the embedding model to be available; without it only exact and alias matches work.
When to use
- Different sessions refer to the same person, tool or company by different names.
- You want a stable id before recording facts about an entity.
- You want to check whether an entity is already known without creating it (create_if_missing: false).
Parameters
| Name | Type | Default | Description |
|---|---|---|---|
mentionrequired | string | — | — |
project | string | "general" | — |
type | string | "person" | Entity type: person, technology, project, company, ... |
threshold | number | 0.85 | Cosine similarity threshold for embedding match. |
create_if_missing | boolean | true | — |
Example
Arguments
{
"mention": "Postgres",
"project": "my-api",
"type": "technology"
}Result shape
{
"canonical_id": 14,
"canonical_name": "PostgreSQL",
"matched_via": "alias",
"confidence": 1,
"is_new": false
}matched_via is one of exact, alias, embedding, created, miss or pronoun. Pronouns return canonical_id -1.
Values are illustrative; the keys follow the server's handler. MCP clients receive the result as JSON text content.
Server description
The description the server sends to your agent in tools/list, captured from the v14.7.0 source:
resolve a mention to its canonical entity within a project+type. Cross-session coreference via name/alias index + embedding cosine. Returns canonical_id, matched_via, and is_new flag. Pronouns return -1.