> For the complete documentation index, see [llms.txt](https://docs.controltheory.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/quickstart.md).

# Quickstart

Fastpath >>>>>

{% hint style="info" %}
**Just want to look around first?** Choose **Explore Sample Data** on first launch to see a preloaded demo workspace with live anomalies and a sample incident. No setup required. When you're ready for real value, follow the steps below to integrate your own project.
{% endhint %}

The shortest path to asking an AI assistant about your logs, metrics and traces: create a source, assign its streams to a workspace, and connect your assistant via MCP. Use the [Dstl8 CLI](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/dstl8-cli.md) to on-board at warp speed — `dstl8 setup` runs the whole flow below as one guided command.

### Terminology

* **Source** — A connection that brings telemetry into Dstl8 (e.g. an AWS account's CloudWatch, a Kubernetes cluster, or an OpenTelemetry endpoint). You can add multiple sources of the same type with different credentials.
* **Signals** — The kinds of telemetry a source carries: logs, metrics and traces. Not every source carries all three.
* **Stream Type** — A kind of stream produced by a source, e.g. a CloudWatch Lambda log group, or a Kubernetes Deployment or DaemonSet.
* **Stream** — A specific instance of a stream type, e.g. one Lambda function or one pod.
* **Service** — An application that sends telemetry, named by its `service.name`. Dstl8 ties a service's logs, metrics and traces together on its [service page](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/services.md).
* **Workspace** — A boundary that controls user and data access. Streams are assigned to workspaces to make them queryable. Every organization starts with a **Default** workspace.

#### Explore with sample data

If you picked **Explore Sample Data** on first launch, you'll notice Dstl8 creates a second workspace tagged **Sample** alongside your **Default** workspace. It comes preloaded so you can see the product working before connecting anything:

* An OpenTelemetry source named **Sample Data** feeding a set of live streams.
* Anomalies and at least one active incident generated from the sample traffic.
* A completed **Möbius Analysis** summarizing the incident, plus the **Sentiment Heatmap** and **Möbius Chat**.

Sample incidents behave like real ones, which means the org admin receives email notifications for them. This is expected. It also previews how alerting works once you connect a real source.

Use the sample workspace to get familiar with the dashboard, incidents, and Möbius. You can delete it anytime from the workspace settings (gear icon) once you have real data flowing.

### 1. Create a source and validate streams

Pick a platform and add it as a source:

* [**AWS CloudWatch**](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/sources/aws-cloudwatch.md) — if your logs live in CloudWatch Log Groups.
* [**Kubernetes**](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/sources/kubernetes.md) — install the in-cluster agent to collect container logs, cluster events and, if you turn them on, metrics.
* [**OpenTelemetry**](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/sources/opentelemetry-otel.md) — send logs, metrics and traces from an OpenTelemetry Collector or SDK. Traces come from an OpenTelemetry source or the Docker agent; from Kubernetes, use an OpenTelemetry source.
* [**Docker**](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/sources/docker.md) — run the agent on a Docker host to collect container logs and forward OpenTelemetry metrics and traces.
* Other sources (Google Cloud, Vercel, Supabase, Cloudflare, Railway, GitHub) — see [Sources](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/sources.md).

New sources show **Listening for telemetry**. Within a couple of minutes of data arriving, the source turns **Healthy** and streams appear with their inferred stream types.

**Validate:** expand the source card on the **Sources** page — you should see streams listed under **STREAMS**. If nothing appears after a few minutes, re-run **Test Connection** from the source's actions menu (gear icon). For AWS CloudWatch, check that the setup stack finished creating in AWS; see [If logs stop arriving](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/sources/aws-cloudwatch.md#if-logs-stop-arriving).

### 2. Assign streams to a workspace

Streams must be assigned to a workspace before they're queryable. Every organization starts with a **Default** workspace — the fastest path is to assign everything to it.

1. On the source card, open the actions menu (gear icon) and select **Assign Streams**.
2. Check the **Default** workspace to assign all streams from this source, or drill in to select specific **Stream Types** or individual streams.
3. Click **Save**.

The source card now shows `1 workspace` receiving streams.

For a source that sends more than one signal, you can also choose which signals (logs, metrics, traces) each workspace gets; see [Signals](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/signals.md). For splitting streams across multiple workspaces (e.g. prod vs staging), see [Assigning streams to workspaces](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/sources.md#assigning-streams-to-workspaces).

### 3. Connect your AI assistant via MCP

Dstl8 exposes an MCP endpoint per organization that lets Claude (or any MCP-compatible assistant) query your logs, metrics, traces, incidents and more directly.

The fastest way is the CLI: `dstl8 install` finds the AI clients on your machine and configures them for you. See [Dstl8 CLI](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/dstl8-cli.md).

To set a client up by hand:

1. Open the **MCP / CLI** page from the left navigation and copy the **MCP Server URL** — it looks like `https://<org_id>.app.dstl8.ai/mcp`.
2. In **Organization Settings → Tokens**, click **Create API Token** and copy the token (it's only shown once).
3. Configure your client. Per-client setup guides:
   * [Claude Code](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/mcp-integration/claude-code)
   * [Claude Desktop](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/mcp-integration/claude-desktop)
   * [Cursor](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/mcp-integration/cursor)
   * [Codex](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/mcp-integration/codex)
   * [ChatGPT](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/mcp-integration/chatgpt)

Full reference: [MCP Integration](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/mcp-integration).

### 4. Ask a question

From your configured assistant, try:

* *"What's going on in production?"*
* *"Any anomalies in the last hour?"*
* *"Summarize recent errors across my streams."*
* *"Which services are slowest right now, and why?"* (with traces)
* *"Is any container about to run out of memory?"* (with metrics)

If you get a grounded answer referencing your actual services and log patterns, you're set.

### 5. Look around

* **Dashboard** — your workspace at a glance: Möbius's analysis, active incidents, and recent activity across your signals.
* **Services** — every service sending telemetry, with its errors, latency, incidents and changes. See [Services](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/services.md).
* **Explore → Metrics** and **Explore → Traces** (beta) — see [Metrics](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/metrics.md) and [Traces](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/traces.md).
* **Incidents** — what Möbius found, with evidence and suggested next steps. See [Incidents](https://docs.controltheory.com/controltheory-documentation/dstl8-docs/incidents.md).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.controltheory.com/controltheory-documentation/dstl8-docs/quickstart.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
