🧠 Redmine Wiki Assistant

A smart AI-powered assistant that connects to your self-hosted Redmine Wiki, indexes the content using OpenAI's Assistants API + Vector Store, and allows you to search, summarize, and ask questions via a simple Streamlit Web UI.

✨ Features

• 🔍 Semantic search across your Redmine wiki
• 💬 Natural language Q&A using GPT-4
• 📄 Displays source wiki pages used in answers
• 🖥️ Clean, easy-to-use web interface
• 🧠 Powered by OpenAI Assistants + Retrieval

⚙️ Setup Instructions

1. 🔧 Clone the Repository

2. 📦 Install Dependencies

If

3. 🛠️ Create

.env

File

🚀 Run the Assistant

Step 1: Load Wiki Content & Create Vector Store

Step 2: Create the OpenAI Assistant (with retrieval)

Step 3: Start the Web UI

💡 Example Usage

Ask questions like:

• "How do I configure email notifications in Redmine?"
• "What are the steps to create a new project?"
• "Summarize the Deployment page."

The assistant will respond based on your wiki — and cite the source content it used!

🧱 Project Structure

.
├── redmine_client.py         # Pull wiki pages from Redmine
├── load_and_embed.py         # Create vector store from wiki content
├── create_assistant.py       # Create OpenAI Assistant with retrieval tool
├── ask.py                    # (Optional) Ask from terminal
├── app.py                    # Streamlit Web UI
├── vector_store_id.txt       # Saved vector store ID
├── assistant_id.txt          # Saved assistant ID
├── .env                      # Your secrets

✅ To Do / Optional Add-ons

• Summarize individual wiki pages
• Chat memory / follow-up Q&A
• Auto-refresh wiki content daily
• User authentication for multi-user access

📄 License

MIT License — free to use, modify, and share.

🙌 Credits

Built using:

• Redmine REST API
• OpenAI Assistants API
• Streamlit