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

# Verify the Origin Agent Is Running and Capturing AI

> Confirm the Origin agent is installed, healthy, and capturing AI activity on macOS and Windows using tray diagnostics, service checks, and test prompts.

After installing the Origin agent, run through this verification checklist to confirm the agent is registered, telemetry collection is configured, and AI activity is flowing into the Origin console. Work through the five sections in order — each one builds on the last.

***

## 1. UI and Diagnostic Report

The Origin agent runs a small UI accessible from the **system tray** (Windows) or **menu bar** (macOS). The icon color is the fastest health check available without opening a terminal.

| Icon color | Status                                                      |
| ---------- | ----------------------------------------------------------- |
| **Green**  | Healthy — agent running, registered, and capturing          |
| **Yellow** | Degraded — running but one or more checks are failing       |
| **Red**    | Error — agent stopped, unregistered, or interception broken |

Click the icon to open the menu. From there you have two key actions:

* **Diagnostic Report** — runs a full self-test covering registration, mTLS credentials, backend reachability, telemetry collection, and entitlements. Every line should read **OK**.
* **Open Log Folder** — opens the runtime log directory directly in Finder or File Explorer:

<Tabs>
  <Tab title="Windows">
    ```text theme={null}
    C:\ProgramData\Origin\Agent\Logs\
    ```
  </Tab>

  <Tab title="macOS">
    ```text theme={null}
    /Library/Origin/logs/
    ```
  </Tab>
</Tabs>

If the icon is green and the Diagnostic Report comes back clean, the agent is healthy locally. Move on to the service check below to confirm both processes are running at the OS level.

***

## 2. Verify services and processes

Both the **Origin Agent** and the **Origin Watchdog** must be running. The watchdog starts at boot and is responsible for bringing up — and keeping up — the core agent process.

<Tabs>
  <Tab title="Windows">
    Run the following command in an **elevated PowerShell** prompt:

    ```powershell theme={null}
    Get-Service OriginWatchdog, OriginAgent
    ```

    Both services should report a status of **Running**.
  </Tab>

  <Tab title="macOS">
    Run the following command in Terminal:

    ```bash theme={null}
    sudo launchctl list | grep origin
    ```

    Two entries should appear: `com.origin.watchdog` and `com.origin.agent`, each with a non-zero PID in the first column.
  </Tab>
</Tabs>

### Cross-check in the Origin Platform

Sign in to the **Origin Platform**, open the **Registered Endpoints** view, and confirm your endpoint appears with a recent **Snapshot Time**. If the endpoint is listed here, it has successfully registered with the Origin Platform.

***

## 3. Confirm telemetry collection is configured

Origin collects AI event data at the application layer — through each AI tool's OpenTelemetry exporter and through application-level hooks — so there is no proxy, system extension, or certificate to verify on either platform.

The agent runs a local OTel collector on the endpoint's loopback interface, on port 4318 by default. Confirm nothing else has claimed that port:

<Tabs>
  <Tab title="Windows">
    ```powershell theme={null}
    Get-NetTCPConnection -LocalPort 4318 -State Listen -ErrorAction SilentlyContinue
    ```
  </Tab>

  <Tab title="macOS">
    ```bash theme={null}
    sudo lsof -nP -iTCP:4318 -sTCP:LISTEN
    ```
  </Tab>
</Tabs>

The listener should belong to the Origin agent.

<Warning>
  If another process already owns the port — a corporate OTel collector such as Datadog, Alloy, or Dynatrace — that host will not capture AI telemetry until the conflict is cleared. Either free the port or set a different **OTLP collector port** fleet-wide under **Settings → Settings**. See [Console Setup](/getting-started/console-setup).
</Warning>

The Diagnostic Report from the tray or menu-bar icon covers this check too, so it is usually the faster route.

***

## 4. Generate test activity

Pick **one or two** of the generation methods below, run a prompt, and then validate the captured activity in the Origin Platform. Choose methods that reflect how AI is actually used in your environment.

<Tip>
  Embed a unique token in every test prompt — for example `ORIGIN-VERIFY-<date>` or a UUID — so you can search for it precisely in the Origin Platform and confirm it was captured. This removes any ambiguity about which activity came from your test.
</Tip>

### Sample prompts

The following prompts exercise different capture archetypes:

* **Summarization** — *"`ORIGIN-VERIFY-20260513` — summarize the NIST AI Risk Management Framework (AI RMF 1.0) in three bullets."*
* **Code generation** — *"`ORIGIN-VERIFY-20260513` — write a Python function that validates an IPv4 CIDR string."*
* **Rewrite / tone shift** — *"`ORIGIN-VERIFY-20260513` — rewrite this for a non-technical executive: \[paste a short paragraph]."*
* **Open Q\&A** — *"`ORIGIN-VERIFY-20260513` — list the OWASP Top 10 for LLM Applications."*
* **Data analysis** — *"`ORIGIN-VERIFY-20260513` — given this CSV: \[paste 5–10 rows], identify outliers and explain your reasoning."*

### 4.1 M365 Copilot — desktop / Office apps

1. Open **Word**, **Excel**, or **PowerPoint** on the verified endpoint.
2. Click the **Copilot** icon in the ribbon to open the Copilot pane.
3. Send a verification prompt from the list above.
4. Send a follow-up message to generate a multi-turn session.

### 4.2 M365 Copilot — web

1. Browse to [https://copilot.microsoft.com](https://copilot.microsoft.com) and sign in.
2. Send a verification prompt.
3. Send a follow-up in the same conversation.

### 4.3 Direct-to-vendor — Claude.ai / ChatGPT / Gemini

Pick whichever service your environment permits: [claude.ai](https://claude.ai), [chatgpt.com](https://chatgpt.com), or [gemini.google.com](https://gemini.google.com).

1. Sign in to the service.
2. Send a verification prompt from the list above.
3. Send two or three follow-up messages to populate a meaningful session.

***

## 5. Verify activity in the Origin Platform

Allow **1–2 minutes** for events to propagate to the platform after generating test activity, then:

1. Open the **Origin Platform**.
2. On the **All Endpoints** view (the default landing page), locate your endpoint and click it.
3. Confirm the **activity bar** at the top of the endpoint page shows recent prompt activity — colored ticks indicate prompts captured within your test window.
4. In the **Resource Overview** graph, confirm that the AI services you used appear as nodes. Running Copilot and Claude.ai traffic should populate at least:
   * **ChatGPT** (M365 Copilot's underlying model surface frequently appears under this label)
   * **Claude**
   * **M365 Copilot**

If the activity bar is empty after a few minutes, re-run the Diagnostic Report from the tray icon and check that every line reads **OK**. Then consult the agent runtime logs for any errors.

<Note>
  If you embedded a unique token in your prompts, use the search or filter functionality in the Origin Platform to find it directly and confirm end-to-end capture.
</Note>