Getting Started

Enagram is a cognitive runtime for AI agents. Each application you create gets its own isolated memory pipeline — processing interaction events into episodic, semantic, procedural, and identity memory in real time.

This guide walks you through creating an application, upgrading it, configuring the memory pipeline, and testing your agent.

---

1. Create an application

After signing in, you will land on the Overview page. If you have no applications yet, use the app switcher in the top navigation bar to create one.

1. Click the app switcher (the stack icon next to your organisation name).
2. Select New application.
3. Enter a name, a URL slug, and choose a region closest to where your agent will run.
4. Click Create — Enagram provisions a Kubernetes StatefulSet, service, ingress, and TLS certificate automatically. This takes around 30 seconds.

Once provisioned, the Overview page shows your application's live memory counts and cluster health.

---

2. Get your API credentials

Navigate to API in the sidebar. You will find:

• API URL — the base URL for all requests to your agent, in the form https://{app-id}.api.enagram.io.
• Deployment ID — used to identify your application in logs and metrics.

All API requests require an Authorization: Bearer {your-api-key} header. You can generate an API key from your Profile page.

Sending your first event

curl -X POST https://{app-id}.api.enagram.io/events \
  -H "Authorization: Bearer {api-key}" \
  -H "Content-Type: application/json" \
  -d '{
    "session_id": "session-001",
    "event_type": "message",
    "payload": {
      "role": "user",
      "content": "I prefer concise answers."
    }
  }'

Events flow through Enagram's 7-layer pipeline and are compiled into long-term memory asynchronously. You can observe this process in the Data panel.

---

3. Upgrade to Pro

By default, applications run on the Free tier — shared compute with modest resource limits. For production workloads, upgrade to Pro for higher CPU and memory limits.

1. Navigate to App Settings → Plan & Usage.
2. Click Upgrade to Pro — $49/mo.
3. You will be redirected to Stripe's hosted checkout. Enter your payment details.
4. On completion, your application tier updates automatically and the resource limits take effect after the next application restart.

You can manage your subscription (cancel, update payment method) at any time via Manage Billing in the same tab.

---

4. Configure the memory pipeline

The Memory Configuration tab in App Settings exposes every tunable parameter in Enagram's cognitive pipeline. Defaults are well-calibrated for general use — only adjust these if you have a specific reason to.

Key parameters to be aware of:

After saving changes, click Save & restart — the application restarts with the new configuration. Changes take effect within ~30 seconds.

Warning: Incorrect values can cause unexpected memory behaviour, data loss, or degraded performance. Changes are applied immediately after restart and cannot be automatically rolled back.

---

5. Test your agent

The Testing tab provides a live chat interface that sends events directly to your application's API and displays the agent's working memory response.

Use it to:

• Verify that events are being stored correctly.
• Check that preferences and facts are being extracted.
• Observe how the memory pipeline responds to corrections and outcomes.

The chat panel sends a POST /events request for each message and then queries GET /working-memory to retrieve the assembled context, displaying it alongside the response.

---

Next steps

• Data — explore raw memory logs, analytics charts, and the memory graph visualiser.
• Observability — monitor salience distributions, procedure success rates, and fact confidence over time.
• API Reference — full endpoint documentation with request/response schemas.