SharePoint to Confluence: Zero-Touch SOP Documentation Pipeline
8-step n8n workflow that takes a single SOP name as input, navigates SharePoint via Microsoft Graph API, filters screenshots by filename pattern, and publishes a structured Confluence page with all images inserted in correct sequence — 90 minutes per SOP to under 2 minutes.
2min
Per SOP (was 90min)
96%
Time reduction
100%
Formatting consistency
0
Documentation backlog
The Problem
- Technical teams capture Greenshot screenshots during implementation (e.g., PROXMOX setup, AnythingML installation) and store them in SharePoint folders.
- Converting those screenshots into formal Confluence SOPs required 1–2 hours manual work per document: download each PNG individually, create the Confluence page, insert images in order, write step descriptions, maintain formatting.
- Documentation was permanently deprioritised.
- Tribal knowledge stayed locked in image files.
- Compliance requirements weren't being met promptly.
- Each team member had different formatting — no consistency across SOPs.
The Solution
8-node n8n workflow. User enters SOP name (e.g., 'PROXMOX') plus config: `{domain, folderName: 'Greenshot', year: '2025', confluenceDomain}`. Auth via Microsoft Entra ID OAuth2.
01SharePoint Discovery
- Get Site Info
- List Drives
- Navigate to Greenshot/2025 folder
- List all PNG files with metadata.
02Intelligent Filtering
- filters files containing SOP name in filename (e.g., 'PROXMOX'), extracts date from filename pattern YYYY-MM-DD-XXXXXX-SOPNAME.png, validates download URLs.
- Example: 2 PROXMOX files from 6 total pass through.
03Batch Download
- Microsoft Graph API downloads each filtered PNG.
04Metadata Extraction
- JavaScript parses filenames into `{pageTitle: 'DRAFTSOP - 2025-02-23 - PROXMOX', date, projectName, totalFiles, filenames[]}`.
05Confluence Page Creation
- creates structured page with H1 project name, date + file count metadata, 3-column SOP table (Step # 100px | Description from filename 400px | Screenshot placeholder 300px).
06Image Upload Loop
- for each screenshot — upload to Confluence as attachment, retrieve current page HTML + version number, replace `[placeholder]` with `<ac:image><ri:attachment ri:filename='filename.png' /></ac:image>` macro, increment version and save.
07Final publication
- all images inserted in sequential order, page version finalised in Confluence.
Results
SOP creation time: 90 minutes manual → under 2 minutes automated — 96–98% reduction per document
100% formatting consistency across every SOP — identical 3-column table structure, correct image sequence, no variation between team members
Screenshots converted to searchable, versioned Confluence documentation immediately after capture — zero backlog
Example: PROXMOX setup SOP (2 steps) fully documented in under 2 minutes; AnythingML installation (4 steps) published as self-service onboarding resource
Audit trail maintained automatically — Confluence version history incremented correctly on every image insertion
Technical staff freed from documentation overhead — implementation work captured at the moment it happens, not days later from memory
Tech Stack
Want something like this?
Tell me what you're building. I'll tell you honestly if I can help.