Memory
Search agent memories and retrieve user profile data.
Overview
The Memory API exposes the SuperMemory-backed agent memory system. It lets you inspect what the AI copilot "remembers" about the user and their workspace interactions -- past decisions, preferences, and context from prior conversations.
Search Memories
GET /api/memory/search
Search user memories by natural language query using semantic search.
Query Parameters
| Name | Type | Required | Description |
|---|---|---|---|
query | string | Yes | Natural language search query (1-500 characters) |
workspace_id | string | No | Optional workspace scope for filtering |
top_k | integer | No | Maximum number of results (default: 5, max: 20) |
Example Request
curl "https://api.praxiomai.xyz/api/memory/search?query=onboarding%20preferences&top_k=5" \
-H "Authorization: Bearer $TOKEN"
import requests
resp = requests.get(
"https://api.praxiomai.xyz/api/memory/search",
headers={"Authorization": f"Bearer {token}"},
params={
"query": "onboarding preferences",
"top_k": 5,
},
)
const resp = await fetch(
"https://api.praxiomai.xyz/api/memory/search?query=onboarding+preferences&top_k=5",
{ headers: { Authorization: `Bearer ${token}` } }
);
Response (200 OK)
{
"memories": [
{
"content": "User prefers data-driven PRDs with quantitative metrics",
"metadata": {
"source": "conversation",
"timestamp": "2026-03-25T14:30:00Z"
},
"score": 0.92
},
{
"content": "User focuses on B2B SaaS onboarding flows",
"metadata": {
"source": "workspace_context",
"timestamp": "2026-03-20T09:15:00Z"
},
"score": 0.85
}
],
"total": 2,
"query": "onboarding preferences"
}
If the memory service is unavailable, returns an empty result set.
Get User Profile
GET /api/memory/profile
Get the user's dynamic profile from SuperMemory. Returns stable user characteristics (static facts) and recently observed patterns (dynamic facts) used by the agent for personalisation.
Example Request
curl https://api.praxiomai.xyz/api/memory/profile \
-H "Authorization: Bearer $TOKEN"
Response (200 OK)
{
"static_facts": [
"Senior PM at a B2B SaaS company",
"Prefers quantitative success metrics",
"Uses Linear for project tracking"
],
"dynamic_facts": [
"Currently focused on reducing churn in onboarding",
"Recently shifted priority to mobile-first features"
],
"is_available": true
}
If the memory service is unavailable:
{
"static_facts": [],
"dynamic_facts": [],
"is_available": false
}Was this helpful?