Confluence Connector - Technical Manual

The Confluence Connector is a Node.js-based service that synchronizes page content and file attachments from Confluence to the Unique platform. It supports both Confluence Cloud and Confluence Data Center deployments.

Core Capabilities:

• Discovers pages via Confluence Query Language (CQL) label searches
• Fetches page content in HTML (Confluence storage representation) and downloads file attachments (including embedded images)
• Computes per-space file diffs against previously ingested state to detect new, updated, and deleted items
• Ingests content into the Unique knowledge base via the Unique API
• Manages scope hierarchies automatically (root scope plus one child scope per Confluence space)
• Operates on a configurable cron schedule (default: every 15 minutes)

Documentation

Key Concepts

Pull-Based Architecture

The Confluence Connector pulls content from Confluence on a scheduled basis:

• The connector queries the Confluence REST API for labeled pages using CQL
• Content is fetched, diffed, and ingested into Unique
• Scheduling is controlled via a per-tenant cron expression (default: */15 * * * *)

Label-Driven Page Discovery

Pages are not automatically synced. Users must apply configurable Confluence labels to mark individual pages or entire page trees for synchronization:

File Diff Mechanism

The connector maintains state to detect changes:

• New items: Labeled pages and attachments not previously synced
• Updated items: Items with changed metadata since last sync
• Deleted items: Previously synced items no longer labeled or deleted from Confluence

Standard References

• Confluence Cloud REST API
• Confluence Data Center REST API
• Confluence Query Language (CQL)