Confluence MCP Server v2.0 - Selective Editing

Confluence MCP Server v2.0 introduces selective editing capabilities for AI assistants.

A MCP (Model Context Protocol) server that enables AI assistants to interact with Confluence. Beyond basic page management, v2.0 introduces targeted editing that allows AI to make specific modifications while preserving existing content, formatting, and macros.

✨ What's New in v2.0

🎯 Targeted Editing Capabilities

v2.0 enables selective modifications instead of full page replacement:

• Section Updates: Modify specific sections under headings without touching other content
• Pattern Replacement: Find and replace text while preserving XML structure and macros
• Table Cell Editing: Update individual table cells without affecting table structure
• Structure Preservation: Maintains all formatting, layouts, macros, and custom elements

🧠 XML-Aware Operations

• Confluence Storage Format Support: Understands and preserves Confluence's XML structure
• Macro Safety: Avoids breaking macros, layouts, or custom elements during edits
• Content Analysis: Identifies safe editing locations and content boundaries
• Rollback Capability: Automatic backup creation for safe editing operations

🛠️ 14 Tools Available

Standard Tools (10): Complete page, space, attachment, and comment management 🆕 Selective Editing Tools (4): Targeted editing capabilities

🚀 Selective Editing Tools

1. update_page_section - Section Replacement

Updates specific sections under headings without affecting surrounding content.

Example: Update project status without touching meeting notes

AI: "Update the 'Project Status' section to show completed milestones"
✅ Updates only that section, preserves all other content

Features:

• Heading detection (H1-H6) with hierarchy support
• Flexible matching: case-sensitive, exact match, or fuzzy search
• Nested heading support for complex document structures
• Preserves all macros, layouts, and formatting outside target section

2. replace_text_pattern - Find & Replace with Structure Preservation

Replaces text patterns throughout pages while preserving document structure.

Example: Update product version across documentation

AI: "Replace all instances of 'v1.2.3' with 'v2.0.0' in the API docs"
✅ Updates version numbers while preserving links, macros, and formatting

Features:

• Content detection distinguishes text from XML markup
• Case sensitivity and whole-word matching options
• Replacement limits for controlled changes
• Preserves macros, links, and formatting elements

3. update_table_cell - Table Cell Editing

Updates specific table cells while maintaining table structure and formatting.

Example: Update project metrics in status tables

AI: "Update the completion percentage in row 2, column 3 to 95%"
✅ Changes only that cell, preserves table formatting and other data

Features:

• Zero-based indexing for precise cell targeting
• Rich HTML content support within cells
• Table structure preservation (borders, styling, headers)
• Targeted modification - only specified cell is changed

4. add_section_at_end - Section Addition

Adds new content sections at the end of pages while preserving existing structure.

Example: Append new meeting notes without affecting existing content

AI: "Add a new section called 'Action Items' with today's tasks at the end"
✅ Adds new section, preserves all existing content and formatting

Features:

• Automatic heading level detection and assignment
• Document structure preservation during insertion
• Safe positioning at document end without content disruption
• Maintains all macros, layouts, and formatting in existing content

🌐 Universal Deployment Platform

🚀 Deployment Options

Platform	Transport	Status	Tools Available
Claude Desktop	stdio	✅ All 14 Tools	Local development, personal use
Smithery.ai	HTTP	✅ All 14 Tools	Cloud deployment, team sharing
Docker	HTTP/stdio	✅ All 14 Tools	Containerized deployment
Web Clients	HTTP	✅ All 14 Tools	Browser-based AI tools
Cloud Platforms	HTTP	✅ All 14 Tools	Railway, Heroku, AWS, etc.

🛠️ Complete Tool Suite

📄 Standard Confluence Tools (10)

Tool	Description	Example Usage
get_confluence_page	Retrieve page content	"Show me the content of page ID 123456"
create_confluence_page	Create new pages	"Create a meeting notes page in the PROJ space"
update_confluence_page	Modify entire pages	"Replace all content in page 123456"
delete_confluence_page	Remove pages	"Delete the outdated page 123456"
search_confluence_pages	Search with CQL	"Find all pages in PROJ space modified this week"
get_confluence_spaces	List available spaces	"What spaces do I have access to?"
get_page_attachments	View page attachments	"Show attachments on page 123456"
add_page_attachment	Upload files	"Upload this document to page 123456"
delete_page_attachment	Remove files	"Delete the old attachment from page 123456"
get_page_comments	Read page comments	"Show me comments on page 123456"

🎯 Selective Editing Tools (4)

Tool	Description	Key Capability
update_page_section	Section replacement	Updates specific headings without touching other sections
replace_text_pattern	Pattern replacement with structure preservation	Replaces text while preserving macros and formatting
update_table_cell	Table cell editing	Modifies individual cells without affecting table structure
add_section_at_end	Section addition	Adds new sections at the end of pages while preserving structure

🚀 Quick Start

📚 New to Selective Editing?

👉 Complete Step-by-Step Tutorial - Learn selective editing with hands-on examples!

Option 1: Claude Desktop (stdio transport)

1. Install Dependencies:

   pip install -r requirements.txt
   

2. Configure Claude Desktop:

   {
     "mcpServers": {
       "confluence": {
         "command": "python",
         "args": ["-m", "confluence_mcp_server.main"],
         "env": {
           "CONFLUENCE_URL": "https://your-org.atlassian.net",
           "CONFLUENCE_USERNAME": "your-email@domain.com",
           "CONFLUENCE_API_TOKEN": "your-api-token"
         }
       }
     }
   }
   

3. Restart Claude Desktop and look for the 🔨 hammer icon showing 14 tools

Option 2: HTTP Server

1. Install Dependencies:

   pip install -r requirements.txt
   

2. Set Environment Variables:

   export CONFLUENCE_URL="https://your-org.atlassian.net"
   export CONFLUENCE_USERNAME="your-email@domain.com"
   export CONFLUENCE_API_TOKEN="your-api-token"
   

3. Start HTTP Server:

   python -m confluence_mcp_server.server_http_optimized
   

4. Test the Server:

   curl http://localhost:8000/mcp
   # Returns: 14 tools including selective editing capabilities
   

Option 3: Docker Deployment

1. Build Container:

   docker build -f Dockerfile.smithery -t confluence-mcp-server-v2 .
   

2. Run Container:

   docker run -p 8000:8000 \
     -e CONFLUENCE_URL="https://your-org.atlassian.net" \
     -e CONFLUENCE_USERNAME="your-email@domain.com" \
     -e CONFLUENCE_API_TOKEN="your-api-token" \
     confluence-mcp-server-v2
   

Option 4: Smithery.ai Deployment

1. Upload Files: Upload smithery.yaml and server code to Smithery.ai
2. Configure Credentials: Set Confluence URL, username, and API token
3. Deploy: Smithery.ai automatically provides all 14 tools including selective editing

💬 AI Assistant Workflows

🎯 Selective Editing Examples

📚 Need Step-by-Step Guidance? 👉 Complete Selective Editing Tutorial - Hands-on guide with examples, workflows, and troubleshooting

Update Project Status Without Disrupting Meeting Notes:

You: Update the project status section to show we completed the API development milestone

AI: I'll update just the project status section while preserving all your meeting notes...
[Uses update_page_section tool]
✅ Updated "Project Status" section only - meeting notes, action items, and formatting preserved

Fix Typos Across Documentation:

You: Fix the typo "recieve" to "receive" throughout our API documentation pages

AI: I'll find and fix that typo while preserving all links and formatting...
[Uses replace_text_pattern tool]  
✅ Fixed 12 instances across 5 pages - all macros, links, and formatting preserved

Update Metrics in Status Dashboard: