Skip to content

Copilot Readiness

An AI feature runs only when five things fit together, and they live in five places: the license of its app, the AI connection, a model it may use, its Copilot capability in Business Central and its own switch in the app. The Copilot Readiness page checks all of this for every AI feature of the merchantCENTRAL apps at once and names, for each feature, the first thing that is missing.

Help in context (F1)

Press F1 to open this page.

Opening the page

  • Search for Copilot Readiness
  • From the AI Setup and the AI Overview, action Copilot Readiness
  • From the setup and the setup wizard of the SEO Engine
  • From a "Copilot is not ready" error, via Open Copilot Readiness

What the page shows

At the top are the checks all features share:

Line What is checked
AI connection Whether endpoint, deployment and API key are entered, and whether the standard model is a reasoning model.
Monthly budget Whether the token or cost budget of the month is reached. All AI features then stop until the first of the next month.

Below, each AI feature has a line, for example Marketplace Text Translation (merchantCENTRAL) or Listing Content Draft, Marketing Text Translation and Image Alt Texts (SEO Engine). The checks run in the order a call meets them:

  1. License of the app
  2. Enabled: the app is switched on in the hub
  3. AI connection
  4. Copilot capability switched on under Copilot & AI capabilities
  5. Model: a deployment the feature may use; for image alt texts an assigned one that reads pictures
  6. Own switch of the feature in its app, for example Draft Alt Texts from Pictures
Status Meaning
Ready The feature can run. Details names the deployment it runs on.
Attention The feature is switched off in its app, or a setting it needs to work is missing. That is often intended.
Missing The feature does not run until the problem named under Details is fixed.

Actions

Action Effect
Fix Opens the page that solves the problem of the selected line: the app's license, the AI Setup or Copilot & AI capabilities. The checks run again when you come back.
Test Connection Asks the standard model a short question. If a reasoning model answers, the test switches Reasoning Model on in the AI Setup.
Check Again Checks every line again.
AI Setup, Copilot & AI Capabilities Open the two pages directly.

merchantCENTRAL switches to reasoning models by itself

GPT-5 and the o-series refuse the parameters Business Central's AI module always sends (max_tokens and a temperature). The first call used to end with a technical Azure message. Now merchantCENTRAL remembers, at the first such error, that the model is a reasoning model and sends the same call directly to Azure OpenAI. The feature stays switchable as a Copilot capability, the budget is checked and the call appears in the usage log.

For developers

An app with an AI feature of its own registers it through the event OnCollectCopilotFeatures of codeunit ALN MC Copilot Readiness: with AddFeature (Copilot capability, connector type and code from the hub's registry) and, if the feature has a switch of its own, SetSwitch with the entry number it returned. License, activation, connection, model and Copilot capability are judged by the hub.

See also