Skip to main content

The SyteOps bridge

The SyteWide MCP plugin is not limited to the Fluent suite REST endpoints. It also carries two tools that speak directly to the SyteOps Manage API, giving an MCP client (Claude, a custom agent, etc.) the ability to inspect and control anything SyteOps manages — without ever leaving the tool-call interface.

How the MCP reaches SyteOps​

Two tools bridge the gap:

MCP toolManage API endpointPurpose
syteops_manifestGET /syteops/v1/manage/manifestDiscover every resource and action SyteOps currently exposes
syteops_dispatchPOST /syteops/v1/manage/dispatchExecute a specific resource/action pair with optional params

The tools call those endpoints on the same WordPress site where the MCP plugin is installed. No additional credentials or network paths are required.

A typical flow​

The intended usage is a two-step pattern: discover, then act.

  1. Call syteops_manifest — it returns the full list of available resources and the actions each one accepts. This lets the AI (or calling code) know what operations are possible before committing to one.
  2. Call syteops_dispatch with the chosen resource, action, and any parameters.

Dispatch call shape​

The body sent to the Manage API by syteops_dispatch looks like this:

{
  "resource": "plugins",
  "action": "activate",
  "params": {
    "slug": "fluent-crm"
  }
}

The API responds with a uniform envelope:

{
  "ok": true,
  "data": { "status": "active" },
  "changed": true,
  "error": null
}
Destructive operations

Actions that are irreversible (deletions, resets, etc.) require an extra confirm: true field in the dispatch body. Omitting it will cause the API to reject the request rather than proceed.

Discover before dispatching

Always call syteops_manifest first in an automated flow. The manifest is the authoritative source of which resource / action pairs are valid — hard-coding them risks calling an action that has been renamed or removed.

Stack diagram​