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

# Memory Kick Start

> Write your first memory and retrieve it with Recallio

Follow these steps to capture personalized memories and surface them in assistant responses.

## Step 1 — Create an API key

1. Sign in to the dashboard and open **API Keys**.
2. Create a key with at least the **Contributor** role. Use a **Manager** key only if you plan to run delete or export operations.
3. Copy the value and store it in your secrets manager. All Memory endpoints require the bearer token.

<Warning>
  Rotate the key immediately if it is exposed. Memory data can include sensitive preferences or history.
</Warning>

## Step 2 — (Optional) Organize memories by project

If you manage multiple environments, go to **Settings → Projects** and create entries that mirror your homes or regions. Each project exposes a `ProjectId`. Include it in every request to keep memories scoped correctly.

## Step 3 — Store a memory

Use `/api/Memory/write` to add a fact about a user. Remember to set `consentFlag` to `true`.

```bash theme={null}
curl https://app.recallio.ai/api/Memory/write \
  -H "Authorization: Bearer $RECALLIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "userId": "resident-991",
    "projectId": "project_abc",
    "content": "Taylor Brooks prefers maintenance updates after 6pm via email.",
    "tags": ["preferences", "communications"],
    "metadata": {"capturedBy": "rpa-bot", "source": "post-visit-form"},
    "consentFlag": true
  }'
```

The endpoint responds with `202 Accepted` while the memory is embedded in the background.

## Step 4 — Recall the memory

Query `/api/Memory/recall` using the same `userId` and `projectId`. Adjust `limit` or `similarityThreshold` as needed.

```bash theme={null}
curl "https://app.recallio.ai/api/Memory/recall?limit=5&similarityThreshold=0.3" \
  -H "Authorization: Bearer $RECALLIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "userId": "resident-991",
    "projectId": "project_abc",
    "scope": "user",
    "query": "communication preferences",
    "tags": ["preferences"]
  }'
```

The response is an array of memories with similarity scores. Feed the top result directly into your assistant prompt.

## Step 5 — Summarize or explore topics (optional)

* Use `/api/Memory/recall-summary` to generate a digest of all facts for the user or project.
* Call `/api/Memory/recall-topics` to see the taxonomy Recallio extracted from the stored memories.

```bash theme={null}
curl https://app.recallio.ai/api/Memory/recall-summary \
  -H "Authorization: Bearer $RECALLIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "userId": "resident-991",
    "projectId": "project_abc",
    "scope": "user"
  }'
```

## Step 6 — Manage data lifecycle (optional)

* **Delete** memories when a resident offboards:

  ```bash theme={null}
  curl -X DELETE https://app.recallio.ai/api/Memory/delete \
    -H "Authorization: Bearer $MANAGER_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "scope": "user",
      "userId": "resident-991",
      "projectId": "project_abc"
    }'
  ```

* **Export** memories for audits:

  ```bash theme={null}
  curl "https://app.recallio.ai/api/Memory/export?Type=fact&Format=json&ProjectId=project_abc" \
    -H "Authorization: Bearer $MANAGER_API_KEY"
  ```

## Next steps

* Browse the [Memory Overview](/memory-overview) for a feature map.
* Dive into the [Memory Deep Dive](/memory-deep-dive) to tune scoring, tagging, and lifecycle policies.
* Consult the [Recallio Swagger reference](https://app.recallio.ai/swagger/index.html) for every field and response model.