Tool reference · Evaluation
memory_eval_temporal
Checks time-aware facts: writes that a test project used one database and later another, then checks both "as of then" and "now" answers.
local stdio server read-only
Caution. Writes two test facts into the real knowledge graph under project "memory_eval_temporal" and does not remove them.
When to use
- You rely on kg_at or temporal facts and want to confirm they work.
- After upgrading, to check the temporal tables exist.
- You want a quick pass/fail for point-in-time queries.
Parameters
| Name | Type | Default | Description |
|---|---|---|---|
limit | integer | — | — |
mode | "fast" | "balanced" | "deep" | "fast" | — |
Example
Arguments
{
"mode": "fast"
}Result shape
{
"scenarios_total": 2,
"scenarios_passed": 2,
"recall_at_5": 1,
"recall_at_10": 1,
"latency_ms": 0,
"mode": "fast",
"llm_calls_during_eval": 0,
"network_calls_during_eval": 0,
"details": {
"query_at_t1": [
"postgres"
],
"current": [
"mysql"
]
}
}If the temporal modules or tables are missing it returns { status: "not_implemented", reason, mode }.
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:
v11.0 Phase 8: temporal recall using temporal_kg + temporal_filter. Returns {status: 'not_implemented', ...} when modules are missing.