TODO: Review and Customize System Prompt
Guides you through reviewing and customizing a system prompt for Llama 4 Maverick to generate hackathon presentations.
What this file does
Guides you through reviewing and customizing a system prompt for Llama 4 Maverick to generate hackathon presentations.
When to use it
- You are preparing a hackathon presentation and need to tailor the AI's output
- You want to align generated slides with specific judging criteria or audience
- You need to coordinate demo automation and avatar timing in the prompt
- You are iterating on prompt quality by testing and refining changes
TODO: Review and Customize System Prompt
๐ Your Task: Review and Elaborate on the System Prompt
The system prompt in config.py is the core instruction that guides Llama 4 Maverick in generating presentations. You should review and customize it based on your specific hackathon requirements.
๐ฏ Current System Prompt Location
File: config.py
Variable: SYSTEM_PROMPT
๐ What to Review and Customize
1. Hackathon-Specific Requirements
- Target Audience: Who are the judges/audience for your hackathon?
- Evaluation Criteria: What are the key criteria for winning?
- Time Constraints: How long do you have for the presentation?
- Technical Focus: What technical aspects should be emphasized?
2. Presentation Style Preferences
- Tone: Formal vs. casual, technical vs. business-focused
- Structure: Preferred presentation flow and sections
- Visual Elements: How should slides/demos be described?
- Engagement: How to keep the audience engaged?
3. Demo Automation Requirements
- Browser Automation: Specific tools or frameworks to mention
- Avatar Integration: How to coordinate with Tavus avatar
- Timing: Synchronization between demo and presentation
- Fallback Scenarios: What to do if automation fails?
4. Technical Depth
- Code Explanation: How detailed should technical explanations be?
- Architecture Focus: Which architectural aspects to highlight?
- Innovation Emphasis: What makes your solution unique?
- Scalability: How to present scalability and future potential?
๐ Suggested Customizations
For Technical Hackathons:
# Add to system prompt:
- Emphasize technical innovation and code quality
- Include code snippets and architecture diagrams
- Focus on performance, scalability, and best practices
- Highlight use of cutting-edge technologies
For Business-Focused Hackathons:
# Add to system prompt:
- Emphasize market opportunity and business model
- Include user stories and customer value
- Focus on revenue potential and competitive advantage
- Highlight team capabilities and execution plan
For Innovation Hackathons:
# Add to system prompt:
- Emphasize creativity and novel approaches
- Include problem-solution fit and market validation
- Focus on disruptive potential and industry impact
- Highlight unique technical or business insights
๐ฌ Demo-Specific Instructions
Consider adding specific instructions for:
- Demo Flow: How to structure the live demonstration
- Error Handling: What to do if things go wrong
- Timing: How to pace the presentation and demo
- Q&A Preparation: Anticipated questions and responses
๐ Next Steps
- Review the current system prompt in
config.py - Identify your hackathon's specific requirements
- Customize the prompt based on your needs
- Test the changes using
test_integration.py - Iterate and refine based on results
๐ Testing Your Changes
After customizing the system prompt:
# Test with a sample repository
python test_integration.py
# Test with the UI
streamlit run demo_ui.py
๐ก Tips for Effective Customization
- Be Specific: Include concrete examples and requirements
- Consider Audience: Adapt language and depth to your judges
- Focus on Impact: Emphasize what makes your solution compelling
- Include Constraints: Mention time limits and technical requirements
- Prepare for Questions: Anticipate what judges might ask
๐ Iteration Process
- Make changes to
SYSTEM_PROMPTinconfig.py - Test with
test_integration.py - Review generated presentations
- Refine the prompt based on results
- Repeat until satisfied
What's inside
6 customization categories, 4 hackathon-type examples, 2 test commands, 5 iteration steps
Change this for your project
- Replace
config.pywith your own configuration file path if different - Replace
test_integration.pywith your own test script name if different - Replace
demo_ui.pywith your own UI entry point if different
Where it goes
Use as the system prompt for your model call, or as the base instructions in your agent framework.
Worth borrowing
- Checklist format for prompt customization ensures no requirement is overlooked
- Separating customization by hackathon type (technical, business, innovation) lets you quickly adapt to different contexts
Related Documents
Character Persona
Defines a Fallout 2 roleplay persona for an LLM agent: charming rogue with social-first combat and moral code.
How to implement tool use
Explains how to define, configure, and control tool use with the Anthropic API, including model selection, tool definitions, output control, and parallel calls.
Gemini by Example
Collects 20+ Gemini API code examples covering text, images, audio, video, PDFs, agents, and token counting.
Claude Tool Use (Function Calling) Documentation
Teaches how to define and use Claude tool use (function calling) with Python, including schema rules, execution patterns, and best practices.