Skip to content

AI Assist

luno includes several AI-powered features that accelerate content creation and CMS configuration — from drafting body copy to suggesting form structures and generating widget styles.

Do this now (3 lines)

  1. Open a text / tiptap field in the entry editor and click AI
  2. Prompt briefly (e.g. “150-char summary”), then accept a suggestion
  3. For agent workflows, go to Path A · Agents

Content AI

The entry editor offers AI-generated suggestions directly inside text and rich-text fields.

Supported field types

Field typeWhat AI can do
textSuggest title variants, generate headlines, propose A/B test options
textareaWrite excerpts, meta descriptions, summaries
tiptapDraft body copy, rewrite sections, translate content

How to use

  1. Open an entry for editing and focus a field
  2. Click the ✨ AI button in the top-right corner of the field
  3. Enter a prompt (or leave it blank for automatic generation based on context)
  4. Review the generated suggestion and click Accept or Regenerate
  5. Manually refine as needed

Example prompts

Title generation for A/B testing:

Prompt: "Suggest 3 title variants for a case study about how a retail company cut content publishing time by 80% using a headless CMS."

→ Output:

1. "From 5 Days to 1 Hour: How Acme Retail Transformed Their Content Pipeline"
2. "80% Faster Publishing: A Retail Chain's Switch to Headless CMS"
3. "Case Study: Scaling Content Operations Without Scaling the Team"

Meta description:

Prompt: "Write a 155-character meta description for a Google Search result based on this article body."

Form Set Field Suggestions

When creating a new form set, describe your content in plain text and the AI will propose a field structure.

How to use

  1. Go to Settings → Form Sets → New Form Set
  2. Click Suggest fields with AI
  3. Describe your content type

Example 1 — E-commerce product catalog:

Prompt: "I'm managing product listings for an online store. I need product name, description, price, stock level, categories, multiple product images, and an SKU."

→ AI suggests:

- name         (text, required)
- description  (tiptap, required)
- price        (number, required, min: 0)
- stock        (number, min: 0)
- categories   (multiselect)
- images       (image)  ← multiple image handling note included
- sku          (text, required)

Example 2 — Technical documentation:

Prompt: "I want to manage API reference documentation with endpoint name, HTTP method, path, request body description, response examples, and version."

→ AI suggests:

- name             (text, required)
- http_method      (select: GET / POST / PUT / PATCH / DELETE)
- path             (text, required)
- description      (tiptap)
- request_body     (tiptap)
- response_example (textarea)
- version          (text)

Widget CSS Generation

Describe your desired widget style in plain language and the AI generates the CSS for you.

How to use

  1. Go to Widgets → Style Settings
  2. Click Generate CSS with AI
  3. Describe the design you want

Example prompt: "Minimal grid layout, 3 columns on desktop, 1 column on mobile. Card style with a subtle border. No shadows. Thumbnail on top, title and date below."

→ Generated CSS:

css
.luno-widget-list {
  display: grid;
  grid-template-columns: repeat(3, 1fr);
  gap: 1.5rem;
}

@media (max-width: 768px) {
  .luno-widget-list {
    grid-template-columns: 1fr;
  }
}

.luno-widget-item {
  border: 1px solid #e2e8f0;
  border-radius: 6px;
  overflow: hidden;
}

.luno-widget-thumbnail {
  width: 100%;
  aspect-ratio: 16 / 9;
  object-fit: cover;
  display: block;
}

.luno-widget-content {
  padding: 1rem;
}

.luno-widget-title {
  font-size: 1rem;
  font-weight: 600;
  margin: 0 0 0.25rem;
  color: #1a202c;
}

.luno-widget-date {
  font-size: 0.8rem;
  color: #718096;
  margin: 0;
}

You can edit the generated CSS in the admin panel editor.

Schema Context (for AI agents)

Go to Settings → Schema Context to add semantic descriptions of your form sets. This helps AI agents understand what your content types mean and how to work with them effectively.

Example schema context

Form Set: blog
Purpose: Technical blog articles for a developer audience.

Fields:
- title: SEO-optimized article headline (60 chars recommended)
- body: Full article body as HTML (tiptap editor). NOT Markdown.
- author: entry_ref to the "authors" form set — use the author's entry UUID.
- tags: Technology slugs (e.g., cloudflare, typescript, cms). Comma-separated.
- og_image: Social sharing image (1200×630 px). Provide as asset UUID.
- meta_description: 155 chars max. Used for search snippets.

This context is available at GET /admin/v1/schema-context and is automatically used by the luno MCP server when an AI agent connects.

AI Agent Integration

luno integrates directly with AI agents via the Model Context Protocol (MCP) and the Agent API.

MCP server setup

json
{
  "mcpServers": {
    "luno": {
      "command": "npx",
      "args": ["-y", "@luno-cms/mcp"],
      "env": {
        "LUNO_API_URL": "https://api.luno.rest/admin",
        "LUNO_AGENT_KEY": "sk-agent-xxxxxxxx"
      }
    }
  }
}

See AI Agents Guide for the full setup.

AI Feature Configuration

To enable AI features, configure an LLM provider in Settings → AI Settings:

SettingDescription
AI ProviderOpenAI or Anthropic
API KeyYour provider's API key
ModelThe model to use (e.g., gpt-4o, claude-3-5-sonnet-20241022)

Cost consideration

AI assist features consume tokens billed by your LLM provider. Usage costs are separate from luno's subscription and depend on how much you use the AI features.

Important Notes

Always review AI-generated content before publishing

AI can produce inaccurate, outdated, or inappropriate content. Use luno's approval workflow (draft → pending_review → published) to ensure human review before any AI-generated content goes live.

  • Generation results vary each time — regenerate for alternatives
  • Very long articles may hit the LLM's context window limit
  • For multilingual content, explicitly specify the target language in your prompt

Next Steps