Hudu Articles

Hudu knowledge base articles: HTML content format, company-scoped vs global articles, article folders (including nesting), drafts vs published, the /api/v1/articles endpoint surface, and search, templ

SKILL.md

GitHub

Hudu Articles Management

Overview

Articles in Hudu serve as the knowledge base, providing a place for runbooks, procedures, network diagrams, SOPs, and general documentation. Articles support rich HTML content, can be organized into folders, and can be scoped to specific companies or kept as global (shared across all companies). MSP technicians rely on articles to quickly find procedures and reference documentation during troubleshooting.

Anti-triggers

  • The same knowledge base in IT Glue — IT Glue calls these Documents; use itglue-documents. Both platforms say "article", "document", and "runbook" interchangeably, so the vendor name is the only signal.
  • A structured record rather than prose — if the thing has fields (make, model, IP, expiry) it belongs on an asset layout, not in article HTML; use hudu-assets.
  • A credential mentioned in a runbook — passwords have their own endpoint and their own audit trail; use hudu-passwords. Do not let an agent paste a credential into article body HTML to "keep it together".
  • A ticket resolution note — work notes belong on the ticket in the PSA, not in the knowledge base; use autotask-ticket-notes-attachments or connectwise-psa-tickets.

Key Concepts

Article Scope

Articles can be scoped in two ways:

Scope Description Use Case
Company-specific Tied to a single company Network diagram for Acme Corp
Global Available across all companies Standard new user setup procedure

Scope is controlled entirely by company_id — omit it (or set it to null) to create a global article.

Article Folders

Folders organize articles within a company or globally, and can be nested via parent_folder_id:

Company: Acme Corporation
+-- Articles
    +-- Procedures
    |   +-- Backup Procedure
    |   +-- Disaster Recovery Plan
    +-- Network
    |   +-- Network Overview
    |   +-- IP Addressing Scheme
    +-- Onboarding
        +-- New User Setup
        +-- Hardware Deployment

Article Content

Article content is stored as HTML. Hudu's editor supports:

  • Headings, paragraphs, lists
  • Tables
  • Images (inline and uploaded)
  • Code blocks
  • Embedded passwords (referenced by ID)
  • Links to other Hudu resources

Draft vs Published

Articles can be saved as drafts before publishing:

State Description
Draft Work in progress, not visible to all users
Published Visible to users with appropriate permissions

Fields

Key fields: id, company_id, name (required), content (HTML), folder_id, draft, slug, created_at, updated_at, url.

See references/fields.md for the complete field reference.

API Patterns

Operation Request
List / filter GET /api/v1/articles?company_id=123&name=backup&page=1
Get one GET /api/v1/articles/456
Create POST /api/v1/articles with { "article": { ... } }
Update PUT /api/v1/articles/456
Delete DELETE /api/v1/articles/456
Archive PUT /api/v1/articles/456/archive
Folders `GET

All requests use the x-api-key header. Request and response bodies are wrapped in a singular resource key (article, folder).

See references/api.md for the complete endpoint catalog with request/response examples.

Common Workflows

Create Comprehensive Runbook

async function createRunbook(companyId, runbookData) {
  // Ensure folder exists
  const folder = await ensureFolder(companyId, runbookData.folderPath);

  // Build content
  let content = `<h1>${runbookData.title}</h1>`;
  content += `<h2>Overview</h2><p>${runbookData.overview}</p>`;

  if (runbookData.prerequisites?.length) {
    content += `<h2>Prerequisites</h2><ul>`;
    content += runbookData.prerequisites.map(p => `<li>${p}</li>`).join('');
    content += `</ul>`;
  }

  if (runbookData.steps?.length) {
    content += `<h2>Procedure</h2><ol>`;
    content += runbookData.steps.map(s => `<li>${s}</li>`).join('');
    content += `</ol>`;
  }

  // Create the article
  return await createArticle({
    name: runbookData.title,
    company_id: companyId,
    folder_id: folder?.id,
    content: content
  });
}

Article Search

The API filters on name only — full-text search across content must be done client-side after fetching.

async function searchArticles(companyId, query) {
  const articles = await fetchArticles({ company_id: companyId });

  const queryLower = query.toLowerCase();
  return articles.filter(article =>
    article.name.toLowerCase().includes(queryLower) ||
    article.content?.toLowerCase().includes(queryLower)
  );
}

Documentation Health Check

async function documentationHealthCheck(companyId) {
  const articles = await fetchArticles({ company_id: companyId });

  const thirtyDaysAgo = new Date();
  thirtyDaysAgo.setDate(thirtyDaysAgo.getDate() - 30);

  const yearAgo = new Date();
  yearAgo.setFullYear(yearAgo.getFullYear() - 1);

  return {
    totalArticles: articles.length,
    drafts: articles.filter(a => a.draft).length,
    recentlyUpdated: articles.filter(a =>
      new Date(a.updated_at) > thirtyDaysAgo
    ).length,
    stale: articles.filter(a =>
      new Date(a.updated_at) < yearAgo
    ).map(a => ({
      name: a.name,
      lastUpdated: a.updated_at
    })),
    empty: articles.filter(a =>
      !a.content || a.content.trim().length < 50
    ).map(a => a.name)
  };
}

Clone Article to Another Company

Folder IDs are company-scoped, so do not carry folder_id across companies — resolve or create a folder in the target company first.

async function cloneArticle(articleId, targetCompanyId, newName) {
  const template = await getArticle(articleId);

  return await createArticle({
    name: newName || template.name,
    company_id: targetCompanyId,
    content: template.content
  });
}

Article Templates

Standard HTML skeletons for Network Overview and Disaster Recovery Plan articles are in references/templates.md.

Gotchas

  • Folder IDs are scoped to a company. Passing a folder_id belonging to another company yields a 422; on failure, retry without folder_id to create at root level.
  • A missing company_id is not an error — it silently creates a global article visible to every company. Double-check before creating client-specific content.
  • Content is raw HTML. Values interpolated into content are not escaped by the API; sanitize any untrusted input before writing.
  • Archive is a distinct verb (PUT /articles/:id/archive), not an archived field on update.

See references/errors.md for the complete error and validation table plus a recovery pattern.

Best Practices

  1. Use consistent structure - Follow templates for standard articles
  2. Organize with folders - Create a logical folder hierarchy per company
  3. Use global articles - Share standard procedures across all companies
  4. Include visual aids - Add diagrams, screenshots, and tables
  5. Include metadata - Add last reviewed date and author at the top
  6. Link related resources - Reference assets and passwords

Related Skills