This guide explains how to create compelling demos, videos, and GIFs for the LLM Red Teaming Platform.
- Quick Demo Scenarios
- Creating Screen Recordings
- Making GIFs
- Demo Script Templates
- Best Practices
- Sample Screenshots
Objective: Show how to run a simple red team scan
Steps:
- Launch application:
streamlit run app.py - Navigate to Configure page
- Select models:
- Target:
gpt-4 - Attacker:
gpt-3.5-turbo
- Target:
- Go to Attack Lab page
- Select vulnerabilities:
- ✅ Robustness
- ✅ Jailbreak
- Set "Attacks Per Vulnerability": 3
- Click Run Red Team Scan
- Show progress indicators
- Navigate to Results page
- Display scan metrics and test cases
- Generate PDF report
Key Points to Highlight:
- Easy configuration
- Real-time progress monitoring
- Clear results presentation
- Professional PDF reports
Objective: Demonstrate custom prompt testing
Steps:
- Navigate to Custom Attack tab
- Select vulnerability type: "Jailbreak"
- Choose jailbreak strategy: "DAN (Do Anything Now)"
- Enter target model:
gpt-4 - Click Execute Custom Attack
- Show real-time response
- Display vulnerability score
- Demonstrate different strategies
Key Points to Highlight:
- Flexible custom testing
- Multiple jailbreak strategies
- Immediate feedback
- Score interpretation
Objective: Show pre-built attack library
Steps:
- Go to Static Attack tab
- Browse attack categories
- Select "Ignore Previous Instructions"
- Run against target model
- Show response and analysis
- Try different prompts from library
Key Points to Highlight:
- Ready-to-use attack prompts
- Organized by category
- Quick testing capability
-
OBS Studio (Free)
- Download: https://obsproject.com/
- Best for: Full-featured recordings
-
ShareX (Free)
- Download: https://getsharex.com/
- Best for: Quick captures
-
Camtasia (Paid)
- Professional editing features
- Best for: Polished demos
-
QuickTime Player (Built-in)
- Screen Recording feature
-
ScreenFlow (Paid)
- Professional quality
- OBS Studio
- SimpleScreenRecorder
- Kazam
Resolution: 1920x1080 (1080p)
Frame Rate: 30 FPS
Bitrate: 5000-8000 kbps
Audio: 128 kbps (if narrating)
Format: MP4 (H.264)
-
Prepare Your Environment
# Clean logs rm logs/*.log # Reset database (optional) rm red_teaming.db python migrate_db.py # Set up demo API keys export OPENAI_API_KEY=demo_key
-
Browser/Window Setup
- Close unnecessary tabs
- Hide personal information
- Use incognito/private mode
- Disable notifications
- Set zoom to 100%
-
Smooth Performance
- Close background applications
- Use a local development environment
- Pre-warm the application
- Have example data ready
- ScreenToGif (Windows) - https://www.screentogif.com/
- GIPHY Capture (macOS) - Free on App Store
- Peek (Linux) - https://github.com/phw/peek
Dimensions: 800x600 or 1024x768
Frame Rate: 15-20 FPS
Duration: 5-15 seconds max
File Size: < 10 MB
Colors: Optimized palette
# Install FFmpeg
# Windows: choco install ffmpeg
# macOS: brew install ffmpeg
# Linux: apt-get install ffmpeg
# Convert with optimization
ffmpeg -i demo.mp4 -vf "fps=15,scale=800:-1:flags=lanczos" \
-c:v gif demo.gif
# Add subtitles
ffmpeg -i demo.mp4 -vf "fps=15,scale=800:-1:flags=lanczos,\
drawtext=text='Running vulnerability scan':fontcolor=white:\
fontsize=24:box=1:boxcolor=black@0.5:boxborderw=5:x=(w-text_w)/2:y=(h-text_h)-50" \
-c:v gif demo_subtitle.gif- Upload video to https://giphy.com/create/gifmaker
- Trim to 5-15 seconds
- Add captions if needed
- Download optimized GIF
-
scan-execution.gif
- Show scan starting and completing
- Display progress bar
- Caption: "Running red team scan in 10 seconds"
-
results-view.gif
- Navigate through results
- Highlight key metrics
- Caption: "View detailed test results"
-
custom-attack.gif
- Execute custom attack
- Show response
- Caption: "Test custom jailbreak prompts"
-
pdf-report.gif
- Generate and preview report
- Caption: "Professional PDF reports"
# LLM Red Teaming Platform Demo
## Opening (30 seconds)
"Hi, I'm going to show you the LLM Red Teaming Platform,
a security testing tool for Large Language Models."
## Problem Statement (30 seconds)
"LLMs are vulnerable to attacks like jailbreaks, prompt
injections, and adversarial prompts. Testing these manually
is time-consuming and inconsistent."
## Solution Overview (1 minute)
"Our platform automates security testing by:
- Generating adversarial attacks
- Testing multiple vulnerability types
- Providing detailed reports
- Supporting all major LLM providers"
## Live Demo (2 minutes)
1. "Let's start by configuring our test..."
2. "I'll select GPT-4 as the target model..."
3. "And choose several vulnerability types..."
4. "Watch as it runs 30 different attacks..."
5. "The results show which attacks succeeded..."
## Results & Reports (1 minute)
"We can see:
- Overall security score
- Individual test results
- Attack prompts and responses
- Professional PDF reports for stakeholders"
## Closing (30 seconds)
"The platform helps security teams identify vulnerabilities
before deployment. It's open-source and available on GitHub."# Technical Deep Dive
## Architecture (2 minutes)
- Show architecture diagram
- Explain component interaction
- Discuss technology stack
## Configuration (2 minutes)
- Environment setup
- API key management
- Provider configuration
## Running Tests (3 minutes)
- Vulnerability selection
- Attack configuration
- Execution and monitoring
## Results Analysis (2 minutes)
- Interpreting scores
- Understanding failures
- Reviewing attack prompts
## Customization (1 minute)
- Custom attacks
- Integration possibilities
- Extension points✅ Do:
- Use high-resolution screen capture
- Maintain consistent window sizing
- Use clear, readable fonts
- Highlight important UI elements
- Add annotations for clarity
❌ Don't:
- Record at odd resolutions
- Show cluttered desktop
- Use tiny fonts
- Include personal data
- Skip setup context
-
Speak Clearly
- Moderate pace (not too fast)
- Avoid filler words ("um", "uh")
- Pause between sections
-
Explain Actions
❌ "I'll click here..." ✅ "I'll click 'Run Scan' to start the vulnerability test..." -
Provide Context
❌ "Now we see the results..." ✅ "The results show 85% of attacks were blocked, indicating good security..."
-
Trim Dead Time
- Remove long loading screens
- Speed up repetitive actions
- Cut mistakes and restarts
-
Add Enhancements
- Highlight mouse cursor
- Add zoom effects for details
- Include text overlays
- Background music (subtle)
-
Professional Touches
- Intro/outro slides
- Transitions between sections
- Consistent branding
- Call-to-action at end
Create these key screenshots for README and documentation:
-
dashboard.png
- Main dashboard with overview metrics
- Show multiple scans in history
-
configuration.png
- Configure page with model selection
- Display multiple providers
-
attack-lab.png
- Attack Lab with vulnerability checkboxes
- Show scan in progress
-
results.png
- Results page with charts
- Display test case details
-
report-preview.png
- PDF report preview
- Show professional formatting
-
custom-attack.png
- Custom attack interface
- Show strategy selection
# Preparing for screenshots
1. Populate with realistic data
2. Use professional-looking model names
3. Include variety in results
4. Show both successes and failures
5. Ensure UI is polishedWindows: Win + Shift + S
macOS: Cmd + Shift + 4
Linux: Shift + PrtScn (or Screenshot tool)
# Install ImageMagick
# Convert to PNG with optimization
convert screenshot.png -resize 1920x1080 -quality 90 optimized.png
# Add border
convert optimized.png -bordercolor "#333333" -border 2 final.png## Demo

### Features in Action
<table>
<tr>
<td><img src="docs/screenshots/dashboard.png" width="400"/></td>
<td><img src="docs/screenshots/results.png" width="400"/></td>
</tr>
<tr>
<td align="center">Dashboard</td>
<td align="center">Results</td>
</tr>
</table>- Upload to YouTube
- Add timestamps in description
- Create playlist for different demos
- Enable closed captions
- Twitter: 2:20 max, optimized for mobile
- LinkedIn: Professional tone, 3-10 minutes
- Reddit: Direct, technical, community-friendly
Before recording:
- Clean environment (no personal data)
- Test run-through
- Prepare script
- Check audio levels
- Close unnecessary applications
- Set appropriate window size
- Have demo data ready
- Test API connectivity
After recording:
- Review for errors
- Add annotations
- Include captions
- Optimize file size
- Test playback on different devices
- Add to documentation
- Share with team for feedback
Last Updated: February 2026
Need help creating demos? Check out the Contributing Guidelines or reach out to the team !