Hands-on tutorials to help you get the most out of KAI-Agent.
Difficulty: Beginner
Time: 15-20 minutes
Learn how to teach KAI-Agent domain-specific knowledge using Skills and Instructions. This tutorial uses a real-world example of analyzing Indonesian cooperative (koperasi) data.
You'll learn:
- How to connect to a database and scan tables
- Why initial queries might fail
- How to create Skills for analysis patterns
- How to add Instructions for business rules
- The iterative learning flow
- Advanced Skills: Creating complex, multi-step analysis skills
- Memory Management: Using long-term memory for user preferences
- Business Glossary: Defining metrics and KPIs
- RAG Integration: Adding document search to your analysis
- MCP Tools: Extending KAI-Agent with external tools
Have a tutorial idea? We welcome contributions! Please:
- Create a new Markdown file in
docs/tutorials/ - Follow the existing tutorial format
- Include clear prerequisites and expected outputs
- Submit a pull request
See CONTRIBUTING.md for guidelines.