Tool reference · Save & recall
memory_explain_search
Runs a fast search and shows how each retrieval tier (keyword, semantic, graph, fuzzy, HyDE) scored the records and how they were merged.
local stdio server read-only
When to use
- A record you expected did not show up and you want to know why.
- You are tuning queries or tags and want to see which tier found what.
Parameters
| Name | Type | Default | Description |
|---|---|---|---|
queryrequired | string | — | — |
project | string | — | — |
type | "decision" | "fact" | "solution" | "lesson" | "convention" | "all" | "all" | — |
limit | integer | 10 | — |
embedding_space | string | array | — | — |
Example
Arguments
{
"query": "billing job lock",
"project": "my-api",
"limit": 5
}Result shape
{
"query": "billing job lock",
"total": 1,
"detail": "full",
"fusion": "rrf",
"total_tokens": 120,
"results": {
"decision": [
{
"id": 1842,
"content": "Use PostgreSQL advisory locks…",
"score": 0.912
}
]
},
"_explain": {
"fts": [
{
"id": 1842,
"bm25": 7.4312
}
],
"semantic": [
{
"id": 1842,
"cos": 0.8123
}
],
"graph": [],
"fuzzy": [],
"hyde": [],
"merged": [
{
"id": 1842,
"score": 0.0328,
"via": [
"fts",
"semantic"
]
}
],
"rerank_applied": false,
"embedding_space": null,
"tiers_used": [
"fts",
"semantic"
]
}
}Explaining a search does not count as a recall, so it does not change the ranking of the records it reports on.
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:
same as memory_search_fast but returns a per-tier breakdown (fts/semantic/graph/fuzzy/hyde with raw scores, the merged RRF list, rerank_applied flag, embedding_space). Use to debug why a record did or didn't surface for a query.