> ## Documentation Index
> Fetch the complete documentation index at: https://docs.memorose.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> The shortest path from a running Memorose node to a real ingestion and retrieval round-trip.

# Quickstart

This page follows the current `memorose-server` API directly.

## What you need

* a running Memorose server on `http://127.0.0.1:3000`
* dashboard credentials
* `jq`
* `uuidgen`

<Note>
  In the current single-node default setup, Memorose auto-initializes Raft at startup. You do not need to call `/v1/cluster/initialize` before sending writes unless you intentionally disabled `raft.auto_initialize`.
</Note>

## The happy path

1. Log in and get a bearer token.
2. Create a stream ID.
3. Ingest one event.
4. Retrieve against that stream.
5. Check whether consolidation backlog is clear.

## 1. Set base variables

```bash theme={null}
export BASE_URL="http://127.0.0.1:3000"
export USER_ID="dylan"
export ORG_ID="default"
```

## 2. Log in and get a token

Current `/v1` routes accept either a dashboard JWT bearer token or an API key. For local development, the dashboard login is the fastest path.

```bash theme={null}
TOKEN=$(curl -s -X POST "$BASE_URL/v1/dashboard/auth/login" \
  -H "Content-Type: application/json" \
  -d '{
    "username": "admin",
    "password": "admin"
  }' | jq -r '.token')
```

## 3. Create a stream ID

```bash theme={null}
STREAM_ID=$(uuidgen | tr '[:upper:]' '[:lower:]')
```

## 4. Ingest an event

```bash theme={null}
curl -s -X POST "$BASE_URL/v1/users/$USER_ID/streams/$STREAM_ID/events" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "org_id": "default",
    "content": "Dylan prefers Rust for systems work and wants concise status updates.",
    "content_type": "text"
  }'
```

Expected response shape:

```json theme={null}
{
  "status": "accepted",
  "event_id": "..."
}
```

## 5. Retrieve memory

```bash theme={null}
curl -s -X POST "$BASE_URL/v1/users/$USER_ID/streams/$STREAM_ID/retrieve" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "org_id": "default",
    "query": "What does Dylan prefer for systems work?",
    "limit": 5,
    "graph_depth": 1
  }'
```

Expected response shape:

```json theme={null}
{
  "stream_id": "...",
  "query": "What does Dylan prefer for systems work?",
  "results": [
    {
      "unit": {
        "id": "...",
        "memory_type": "factual",
        "content": "...",
        "keywords": ["rust", "systems"],
        "level": 1
      },
      "score": 0.9
    }
  ],
  "query_time_ms": 12
}
```

Important response fields:

* `results[].unit.level`: where the retrieved memory sits in the hierarchy
* `results[].unit.memory_type`: `factual` or `procedural`
* `score`: final ranking score after retrieval fusion
* `query_time_ms`: observed server-side latency for the retrieval call

## 6. Check consolidation backlog

```bash theme={null}
curl -s "$BASE_URL/v1/status/pending" \
  -H "Authorization: Bearer $TOKEN"
```

When `"ready": true`, the current pending-event backlog is clear.

## Where to go next

* [Configuration](/getting-started/configuration)
* [Basic Operations](/guides/basic-operations)
* [Cluster API](/api/cluster)
* [Tasks And Goals](/guides/tasks-and-goals)
* [REST API](/api/rest-api)