---
title: "corestack_site_operation_get"
url: https://docs.corefinity.com/mcp/tools/corestack_site_operation_get
type: tool
last_updated: 2026-11-13
feature: operations
related:
  - https://docs.corefinity.com/api/reference/sites/get-operation.md
  - https://docs.corefinity.com/api/long-running-work.md
---
# corestack_site_operation_get

**An operation.** One change made to a site or environment — its type, status, steps, error and result — read from the control panel, which moves it on first if it is due. Scope: read.

- Scope: `read`
- Runs as a task: no

## When to use it

One change made to a site or environment — its type, status, steps, error and result — read from the Control Panel, which moves it on first if it is due. A step that runs a command on the environment is never run inside this call: the Control Panel runs it in the background while you keep reading, so read again to see it progress.

A Varnish VCL change (`varnish_vcl_change`, from `corestack_varnish_vcl_update` or `corestack_varnish_vcl_revert`) also carries `vcl`: `{revision, status, outcome, applied, applied_at, reason, warnings}`. `outcome` is `applied` once Varnish confirmed the new VCL; `pending_verify` (committed, not confirmed yet), `refused`, `verify_failed` (the previous revision is being put back) or `rolled_back` otherwise, with the `reason`. A completed operation is only applied when `vcl.applied` is true.

Every operation carries `notes`: `[{code, text}]`, what it did differently on purpose, in the words the Control Panel shows (`[]` when nothing). `theme_compile_deferred`: a files-only Shopware copy onto an environment whose database holds no shop yet did not compile the theme; copy the database next and that copy compiles it. `php_settings_clamped`: a new environment could not take the memory limit, upload size or max execution time of the environment it was made from as they were; the text names each setting changed to fit its limits and the value it has now. `platform_address_changed`: a new site's or environment's usual platform address (`<domain>-<company>.cfstack.io`) was already used by another site, so it was created on the next free one; the text names the address it uses. `certificate_on_its_way`: a new site or environment is waiting while its certificate is issued ("Certificate on its way: …"); it is not a failure and needs nothing from you: the create carries on by itself, for up to a few hours, and reaches `ready` once the certificate's issuer is ready. The note goes once it has. `search_carried`, `search_stays_off` and `search_not_carried`: a new Shopware environment started with search on because the environment it was made from has it on; kept it off because its size is under Medium (8 GB) although the source has it on; or started with it off because the source's setting could not be read (switch it on from the Shopware tab). A site or environment create that stops on a temporary problem is tried again automatically, at most three times, a few minutes apart: `auto_retry_scheduled` on the failed create (the text says roughly when, and which attempt), `auto_retried` once it was (the text names the new operation, which you can follow instead), `auto_retry_attempt` on the automatic attempt itself (its `requested_by` reads "Corefinity (automatic)"), `auto_retry_exhausted` once every automatic attempt was spent and the create is left to Corefinity's staff, and `create_still_trying` on a create still being tried before it is saved to the platform. Do not create the site again while one of these says it is being tried.

A Shopware copy that kept shop domains naming another site or environment (neither the copy's source nor its target, such as another site's address an earlier import brought in) carries `foreign_domains_kept`; the text says how many. They are left as they were: requests for them still reach that other site, and only the links and sitemaps the shop makes with them lead there. corestack_site_storefront_domains_list lists the shop's domains; change them in the sales channel's domains in Shopware if they should not.

A clone, a copy made to seed a new environment, a move or a CoreStack Enterprise copy that left out files or folders the website cannot read carries `unreadable_paths_skipped`; the text names them ("2 files or folders the website cannot read were not copied: …"). Those paths are not on the copy, so anything that read them on the source does not find them there; make them readable on the source and copy again to bring them across.

A create (`provision`, `env_create`, `cluster_create`) accepted while another change in the same region was starting stays `pending` with the `status_label` "Waiting for another change to finish. This carries on by itself." until it starts; nothing needs to be sent again.

While Corefinity has switched site commands off, a change whose check reads the site (a Node.js or PHP runtime change, error pages, a Varnish VCL) waits rather than timing out: its `status` stays `rolling_out` and its `status_label` is "Waiting: commands are switched off". It is checked and finished once commands are back; nothing needs to be done again.

When the platform needs more capacity for a new site or environment, its create (`provision` or `env_create`) waits while that capacity is prepared instead of failing: its `status` stays `rolling_out` and its `status_label` is "Setting up your site. This can take a little longer." ("…your environment…" for an environment). It carries on by itself once there is room; nothing needs to be done. If it still has no room after about two hours it fails with `failure_reason` `no_room`, a temporary problem, so it is tried again automatically (`auto_retry_scheduled`, above). A task following the create (from `corestack_site_create` or `corestack_environment_create`) reads the same words as its status while it waits.

A site or environment create whose DNS records cannot be published yet (Cloudflare busy, or a platform DNS problem for Corefinity to fix) waits at its `dns` step instead of failing: its `status` stays `pending` and its `status_label` is "Waiting for DNS". It is tried again every minute and goes on by itself; do not create it again.

A create whose platform is not taking new changes for a while (a sync window, or automated sync switched off) waits before it commits anything: its `status` stays `pending` and its `status_label` is "Waiting for the platform". It carries on by itself once the platform accepts changes; do not create it again.

A failed operation carries `failure_reason`, the same code the Control Panel and the REST API show, and `error` in their words. `storage_unavailable`: the region's storage was not ready when the site's first server was prepared ("Storage in this region is not ready for a new site at the moment. Contact support."); do not retry: ask Corefinity support.

## Inputs

- `operation_id` (integer, ≥ 1, required)

### Input schema

```json
{
  "additionalProperties": false,
  "properties": {
    "operation_id": {
      "minimum": 1,
      "type": "integer"
    }
  },
  "required": [
    "operation_id"
  ],
  "type": "object"
}
```

## Refusals

- `wrong_row`: the operation belongs to a cluster, not a site; read it with corecluster_operation_get.
- The refusals every tool can answer: `invalid_arguments`, `tenant_mismatch`, `scope_required`, `pin_refused`, `rate_limited`, `temporarily_unavailable`, `internal_error` (corestack://guides/how-this-server-works).

## Example

```json
{
  "name": "corestack_site_operation_get",
  "arguments": {
    "operation_id": 77120
  }
}
```

## Prompt

```text
Using the Corefinity MCP server, call corestack_site_operation_get with operation_id 77120. (An operation.) Tell me what it returned.
```

## The same task elsewhere

- Endpoint: [GET /api/corestack/operations/{operation}](https://docs.corefinity.com/api/reference/sites/get-operation.md)
- Guide: [Long-running work](https://docs.corefinity.com/api/long-running-work.md)
