Confluence Knowledge Base MCP Server
Turns Confluence documentation into an AI-powered knowledge base, enabling natural language questions about your systems with answers retrieved from your actual documentation through semantic search.
README
Confluence Knowledge Base MCP Server
An MCP server that turns your Confluence documentation into an AI-powered knowledge base for Gemini CLI. Ask natural language questions about your systems and get answers from your actual documentation.
Quick Start
One-Command Setup
git clone <this-repo>
cd confluence-knowledge-base
./install.sh
The interactive wizard will:
- ✅ Install dependencies in a virtual environment
- ✅ Ask for your Confluence credentials
- ✅ Discover your spaces
- ✅ Help you choose which spaces to index
- ✅ Build the initial knowledge base
- ✅ Configure Gemini CLI automatically (merges with existing config)
What You'll Need
Before running the installer:
-
Confluence API Token
- Go to: https://id.atlassian.com/manage-profile/security/api-tokens
- Click "Create API token"
- Copy the token (you won't see it again!)
-
Your Confluence URL
- Example:
https://yourcompany.atlassian.net
- Example:
-
Python 3.8+ installed
- The installer creates a virtual environment automatically (no system-wide packages needed)
-
Gemini CLI installed
- Install from: https://github.com/google-gemini/gemini-cli
Usage
Once installed, just start Gemini CLI and ask questions:
gemini
> How does our authentication system work?
> What's the process for deploying to production?
> Explain our database migration strategy
> What are the API rate limits?
Gemini will automatically retrieve relevant documentation and answer your questions!
How It Works
1. Your Confluence docs → Downloaded and indexed (one-time)
2. You ask a question → Semantic search finds relevant chunks
3. Gemini gets context → Answers based on YOUR docs
Technologies Used
- FastMCP - MCP server framework
- ChromaDB - Local vector database
- sentence-transformers - Semantic search
- Confluence REST API - Documentation retrieval
Project Structure
confluence-knowledge-base/
├── install.sh # Interactive setup wizard
├── confluence_knowledge_base.py # Main MCP server
├── confluence_kb_with_staleness.py # Version with auto-reindex
├── find_space_keys.py # Space discovery utility
├── requirements.txt # Python dependencies
├── KNOWLEDGE_BASE_SETUP.md # Detailed setup guide
└── README.md # This file
Configuration
After installation, configuration is stored in:
- Credentials:
~/.confluence_mcp.env - Index:
~/.confluence_mcp/index/ - Gemini Config:
~/.gemini/settings.json - Virtual Environment:
./venv/(in the project directory)
Updating Documentation
When your Confluence docs are updated:
Option 1: Ask Gemini
> Reindex the Confluence documentation
Option 2: Command line
./venv/bin/python confluence_knowledge_base.py
Option 3: Automated (Weekly)
Set up a cron job (see REINDEXING_GUIDE.md)
Customization
Change indexed spaces
Edit ~/.confluence_mcp.env:
export CONFLUENCE_SPACES="ENG,DEVOPS,TEAM"
Then rebuild the index.
Adjust chunk size
In confluence_knowledge_base.py:
CHUNK_SIZE = 1000 # Default: 1000 characters
CHUNK_OVERLAP = 200 # Default: 200 characters
Change embedding model
For better quality (slower, larger):
self.embedding_model = SentenceTransformer('all-mpnet-base-v2')
Troubleshooting
"Connection failed"
Check that:
- Your Confluence URL is correct
- Your API token is valid
- You have internet connectivity
"No spaces found"
You might not have access to any Confluence spaces. Ask your admin for access.
Slow indexing
Normal for large documentation sets (500+ pages). Reduce spaces or run overnight.
Wrong/outdated answers
Your index is cached! Reindex when docs are updated:
rm -rf ~/.confluence_mcp
./install.sh
Advanced Usage
Manual space discovery
source ~/.confluence_mcp.env
python3 find_space_keys.py
Staleness detection
Use the enhanced version with automatic staleness warnings:
# In ~/.gemini/settings.json, change the args to:
"args": ["confluence_kb_with_staleness.py"]
Add environment variables:
export MAX_INDEX_AGE_DAYS=7
export AUTO_REINDEX=true
Scheduled reindexing
See REINDEXING_GUIDE.md for cron job setup.
FAQ
Q: Does this modify my Confluence documentation? A: No, it's read-only. It only downloads and indexes content.
Q: Where is my data stored?
A: Locally in ~/.confluence_mcp/index/. Nothing is sent to external services except Gemini API calls.
Q: How much does it cost? A: The MCP server is free. You only pay for Gemini API usage (queries to the AI).
Q: Can I use this with Claude instead of Gemini? A: Yes! MCP is a standard protocol. Just configure Claude Desktop to use this MCP server.
Q: How often should I reindex? A: Depends on how often your docs are updated. Weekly is common. Daily if very active.
Q: Can I exclude certain pages?
A: Not by default, but you can modify confluence_knowledge_base.py to filter by title, label, etc.
Q: What about attachments/PDFs? A: Currently only page content is indexed. Attachments could be added with additional code.
Documentation
KNOWLEDGE_BASE_SETUP.md- Comprehensive setup guideREINDEXING_GUIDE.md- Strategies for keeping docs fresh
Contributing
Feel free to:
- Add features (write capabilities, attachment support, etc.)
- Improve chunking strategies
- Add better error handling
- Create additional tools
License
[Your license here]
Support
For issues or questions:
- Check the troubleshooting section above
- Review the detailed guides in
/docs - Open an issue on GitHub
Ready to get started? Just run:
./install.sh
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.