# Track and wait for operations

> Wait for work to finish, find an operation, and retry a create request

> Nevia is in beta and open to selected teams. Public launch in Dec 2026. The CLI installs with `curl -fsSL https://get.nevia.cloud | sh`, and the MCP server is at https://api.nevia.cloud/mcp. A person signs in once in the browser.
> Setup, sign-in and requesting access: [SKILL.md](https://nevia.cloud/SKILL.md). When Nevia fits and when it doesn't: [why.md](https://nevia.cloud/why.md).

> For the complete documentation index, see [llms.txt](https://nevia.cloud/llms.txt). This page's HTML version: https://nevia.cloud/docs/guides/operations

An operation records work such as creating, branching or deleting a resource.
Use its ID to follow the result without starting the work again.

You need the [CLI installed and signed in](https://nevia.cloud/docs/get-started/quickstart.md) and a
[workspace selected](https://nevia.cloud/docs/workspaces/select.md) where you have the admin or editor
role. The name `demo-ops` must be free in that workspace.

## Wait with the command

Create a computer with the platform defaults. Pick an unused idempotency key and
keep it if you need to retry this request:

```bash
nevia new demo-ops --idempotency-key demo-ops-create-1 --wait
nevia status demo-ops
```

`--wait` follows the operation until it finishes or fails. On success, status
shows `ready` with no pending transition. See
[Create a computer](https://nevia.cloud/docs/computers/create.md) for image and configuration choices.

If the wait times out, the operation keeps running. Continue waiting by ID.

## Find and wait for an operation

List recent operations in the selected workspace:

```bash
nevia ops
```

Find the `computer.create` row whose `RESOURCE` matches the computer ID that
`nevia status demo-ops` prints. The row shows the operation ID and state. To list
only running work:

```bash
nevia ops -f state=running
```

Copy the ID from the `OPERATION` column. Replace `01d4cnjm89yk0r3xtv80dy0x7by` below with that ID:

```bash
nevia wait 01d4cnjm89yk0r3xtv80dy0x7by --timeout 10m
```

On success, the command reports `finished`. Check the computer again with
`nevia status demo-ops`. If the operation failed, read the reported error and the
computer's current state before you decide what to do.

You can also wait on an operation ID from an earlier command or from an
`operation_in_progress` error. Waiting again after a timeout picks up the same
work and doesn't restart a failed operation.

## Retry a create after losing the response

Run `nevia status demo-ops` and `nevia ops` first. If you have an operation ID,
wait for it.

To repeat the create request, send the same key, name, workspace and options as
the original:

```bash
nevia new demo-ops --idempotency-key demo-ops-create-1 --wait
```

Within 24 hours, the API replays a stored response by returning the original
computer and operation. Send retries one at a time, because simultaneous requests
with the same key are not combined. A new create needs a new key. See
[Operations and idempotency](https://nevia.cloud/docs/reference/operations.md) for replay limits and
which actions support it.

## Clean up

Once creation has finished, delete the example computer and its data:

```bash
nevia rm demo-ops --yes --wait
```

The command reports that the computer was deleted.

For filters, see [`nevia ops`](https://nevia.cloud/docs/reference/cli/ops.md). For wait options, see
[`nevia wait`](https://nevia.cloud/docs/reference/cli/wait.md). The same approach works when
[creating a database](https://nevia.cloud/docs/databases/create.md).