Skip to Content
Gavana Help Article Standard Format

Gavana Help Article Standard Format

Every article MUST follow this MDX format exactly:

---
title: "How to [Action] in Gavana"
description: "Learn how to [action] [context] in Gavana."
---
 
# How to [Action]
 
[1-2 sentence overview of what this guide covers and why you'd want to do it.]
 
## Prerequisites
 
- [What the user needs before starting, e.g., "An open canvas" or "An existing saved asset"]
 
## Steps
 
### Step 1: [Action verb] [Object]
 
[1-2 sentences explaining this step.]
 
{/* screenshot: [Description of what the screenshot would show] */}
 
### Step 2: [Action verb] [Object]
 
[1-2 sentences explaining this step.]
 
{/* screenshot: [Description of what the screenshot would show] */}
 
[Continue for all steps...]
 
## Tips
 
- [Helpful tip 1]
- [Helpful tip 2]
 
## Related Articles
 
- [Link to related article 1](/section/article)
- [Link to related article 2](/section/article)

Rules:

  1. Title format: “How to [Verb]…” — task-oriented
  2. 5-12 steps per article (not too few, not too many)
  3. Every step has a screenshot callout comment
  4. Always include Prerequisites (if any), Tips, and Related Articles
  5. Use simple language — canvas users, not developers (the agent-access category is the exception: its audience is technical)
  6. Mention specific UI elements by name in bold (buttons, tabs, fields) — verify the literal label text in the component source before writing it, don’t guess
  7. Each category’s _meta.js orders pages logically, and every article file must be registered there or it won’t appear in navigation

Screenshot Placeholder Formats

Simple (default for manually written articles):

{/* screenshot: Description of what the screenshot shows */}

Enriched (auto-generated by Helpdesk Generate workflow):

{/* screenshot: Description | page: /app-route | element: .css-selector | annotate: yes|no|auto */}
![Alt text](/images/article-slug/step-N.png)

Both formats are processed by the ScreenCapture skill’s FromPlaceholders workflow. The simple format relies on smart inference; the enriched format provides explicit instructions.

Common Issues