Knowledge Base

The knowledge base is a collection of articles that serve two purposes: providing self-service content for visitors and grounding AI responses with citations. Articles are organized into categories, support full-text search, and can be published publicly or kept as drafts.

Overview

Knowledge base articles are stored per-organization and support:

• Rich text content with Markdown formatting
• Tags for cross-category discovery
• Slug-based URLs for public access
• Soft delete (articles are never permanently removed)
• Full-text search across titles, content, and tags

Articles have two statuses: draft and published. Only published articles are visible to visitors and used in AI responses.

Creating Articles

Create articles through the dashboard or via the API. A slug is automatically generated from the title:

POST /api/kb/articles

{
  "title": "How to Reset Your Password",
  "content": "To reset your password, go to the login page and click 'Forgot Password'. Enter your email address and follow the confirmation link.",
  "categoryId": "a1b2c3d4-...",
  "tags": ["account", "password", "security"],
  "status": "published"
}

The slug is derived from the title using slugify logic. If a slug already exists, a numeric suffix is appended (e.g., how-to-reset-1).

Categories

Organize articles into categories with names, descriptions, and emoji icons. Categories help operators browse content and help visitors navigate the knowledge base.

Create a Category

POST /api/kb/categories

{
  "name": "Account Management",
  "description": "Articles about account settings, passwords, and profile",
  "icon": "👤"
}

List Categories with Counts

GET /api/kb/collections

Returns categories sorted by published article count, with pagination:

{
  "items": [
    {
      "id": "...",
      "name": "Account Management",
      "description": "Articles about account settings",
      "icon": "👤",
      "articleCount": 12
    }
  ],
  "total": 5,
  "limit": 20,
  "offset": 0
}

AI Integration

Published knowledge base articles are automatically available to the AI chatbot. When a visitor asks a question, the system searches the knowledge base for relevant articles and uses them as context for the AI response, including citations back to the source article.

Public Search Endpoint

The search endpoint is available without authentication for widget and visitor use:

GET /api/kb/search?q=password+reset&orgSlug=my-company&limit=5

{
  "results": [
    {
      "id": "...",
      "title": "How to Reset Your Password",
      "slug": "how-to-reset-your-password",
      "content": "To reset your password, go to the login page...",
      "categoryId": "..."
    }
  ]
}

Get Article by Slug

Retrieve a specific published article by its slug for public display:

GET /api/kb/articles/slug/:slug?orgSlug=my-company

Listing Articles

List articles with optional filters and pagination:

GET /api/kb/articles?search=security&categoryId=a1b2c3d4&status=published&limit=20&offset=0

Query Parameters

Results are sorted by most recently updated.

Public Access

Knowledge base articles can be accessed publicly by slug. This allows you to embed a help center or FAQ page on your website. Public endpoints do not require authentication but need the orgSlug parameter to identify your organization.

Public endpoints:

• GET /api/kb/search: search published articles
• GET /api/kb/articles/slug/:slug: get article by slug

API Reference

Articles (Authenticated)

Categories (Authenticated)

Public (No Auth Required)

Related

• Chatbot Rules Engine: automated rules that use knowledge base content
• API Authentication: authenticating API requests