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

# Operate in Control Center

> Trigger runs, watch status, and fetch outputs from the dashboard UI.

Most production users drive Sequentum through the [REST API](/api-reference/introduction) or [MCP](/mcp/overview), but the Control Center web UI at [`dashboard.sequentum.com`](https://dashboard.sequentum.com) is also a first-class operating surface — handy for debugging a misbehaving agent, triggering a one-off run, or scanning recent failures without writing code.

## Find your runs

Every Run an agent fires lands in **Run history**, filterable by agent, status, and date. Statuses follow the [Run lifecycle](/concepts/runs) — `Success` and `Completed` are green-banded, `Failure` (which is restartable in place) is amber, `Failed` (which is history-only) is red.

## Trigger a run

From any agent's details screen, click **Run** to kick off a new Run. A side panel collects the same parameters the [`POST /agent/{agentId}/start`](/api-reference/agents/start-run) endpoint accepts — input parameters, parallelism, proxy pool, log level. Click **Start** and the new Run appears in the active runs table within a few seconds.

## Stop or kill a Run

Click any active Run to open its detail view.

* **Stop** maps to [`POST /agent/{agentId}/run/{runId}/stop`](/api-reference/agents/stop-run) — graceful shutdown that persists what's already been extracted.
* **Kill** maps to [`POST /agent/{agentId}/run/{runId}/kill`](/api-reference/agents/kill-run) — forceful termination that may discard in-flight work.

Prefer Stop. Use Kill only when Stop isn't responding. See the [Run lifecycle](/concepts/runs) for the full distinction.

## Inspect outputs

The Run detail view has three tabs:

* **Records** — the extracted data, paginated and filterable. Click any record to see the full JSON.
* **Files** — raw HTML, screenshots, exported files, log files. Each downloadable.
* **Diagnostics** — per-command timing, failure traces, blocked-page detection.

For programmatic access to the same data, see [`GET /run/{runId}/files`](/api-reference/agents/run-files) and [`GET /run/{runId}/file/{fileId}/download`](/api-reference/agents/run-file-download).

<Tip>
  For high-volume integrations, configure an [Export](/agents/commands/overview) destination (Snowflake, S3, Webhooks) so records land in your warehouse or queue without polling the UI.
</Tip>