Skip to content

AI Provider Setup

Summary

SkyCMS AI covers more than Monaco inline completions. The current tenant AI configuration powers:

  • Monaco inline completions,
  • Monaco AI chat,
  • CKEditor Visual Editor AI writing assistance,
  • standalone AI Help Chat.

The feature is off by default and requires an administrator to configure the tenant AI provider before it becomes available to editors.

Outcome

After completing this guide, tenant AI should be enabled with a valid endpoint and credentials, and editor-facing AI features should be available where supported.

Where to configure

Open Settings -> AI Provider in the SkyCMS administration panel.

The tenant-level provider settings apply to all editors in that tenant. When the provider supports it, editors can still choose their own preferred model inside Monaco, CKEditor, or AI Help Chat.

Fields

Enabled

Turns AI on or off for the tenant.

Endpoint

The full URL of an OpenAI-compatible chat completions API endpoint.

Examples:

Provider Endpoint
GitHub Models https://models.inference.ai.azure.com/chat/completions
OpenAI https://api.openai.com/v1/chat/completions
Azure OpenAI https://<resource>.openai.azure.com/openai/deployments/<deployment>/chat/completions?api-version=2024-10-21
Local (Ollama) http://localhost:11434/v1/chat/completions

Model

Use auto when the provider supports SkyCMS-managed defaults.

Current behavior:

  • GitHub Models: auto resolves to gpt-4o-mini
  • OpenAI: auto resolves to gpt-4o-mini
  • Azure OpenAI: SkyCMS typically infers the deployment from the configured endpoint

Access Token

The secret sent to the upstream provider.

Current behavior:

  • GitHub Models and OpenAI use bearer-token style auth,
  • Azure OpenAI uses the api-key header.

Timeout (ms)

How long SkyCMS waits for the upstream provider.

Temperature

Controls randomness. Lower values usually work better for editing and code tasks.

Discovered models

The settings page can preview provider model options before you save.

Current SkyCMS behavior:

  • GitHub Models: live model catalog supported
  • OpenAI: live model catalog supported
  • Azure OpenAI: deployment inferred from the endpoint
  • local servers and some advanced gateways: manual model entry may still be required

Max Tokens

Limits how much text the provider can return in one response.

Getting provider credentials

GitHub Models token

  1. Sign in at github.com.
  2. Open Settings -> Developer settings -> Personal access tokens -> Fine-grained tokens.
  3. Generate a new token.
  4. Grant Models: Read permission.
  5. Paste the token into Access Token in SkyCMS AI Provider settings.

Official references:

OpenAI API key

  1. Sign in to your OpenAI platform account.
  2. Create a secret key from the API key page.
  3. Paste it into Access Token.
  4. Use the OpenAI endpoint and leave Model on auto unless you want to pin a specific model.

Official references:

Azure OpenAI API key

  1. Open your Azure OpenAI resource.
  2. Copy the API key and endpoint.
  3. Use the deployment-scoped chat completions endpoint in SkyCMS.
  4. Keep Model on auto unless you need to pin a specific deployment name.

Official references:

Expected behavior after setup

Once settings are saved with Enabled on and a valid endpoint/token:

  1. Monaco can enable inline completions and AI chat.
  2. CKEditor can enable its region-scoped AI assistant.
  3. AI Help Chat can open from the editor menu.
  4. If supported, model pickers can load live model options.

Verification

Configuration is successful when AI status is enabled and configured, editor surfaces can invoke AI actions, and provider-backed responses are returned without authentication or endpoint errors.

For the fuller provider guidance, use the dedicated configuration pages: