examples

Reflexive Python SDK - Examples

This directory contains example applications demonstrating various use cases of the Reflexive Python SDK.

Running Examples

Method 1: With Full CLI Integration (Recommended)

# Install Reflexive CLI (Node.js)
npm install -g reflexive

# Run any example with full AI capabilities
reflexive --debug simple_app.py
reflexive --debug web_server.py
reflexive --debug data_pipeline.py

This gives you:

• ✅ Full .chat() functionality
• ✅ Debugger with breakpoints
• ✅ Web dashboard
• ✅ Real-time log streaming
• ✅ State inspection

Method 2: Standalone

# Install Python SDK
pip install reflexive

# Run standalone (limited chat capabilities)
python simple_app.py
python web_server.py
python data_pipeline.py

Examples

1. simple_app.py - Basic Usage

What it demonstrates:

• Creating a Reflexive instance
• Setting and tracking state
• Logging events
• Using .chat() to ask AI questions

Run it:

reflexive --debug simple_app.py

What to try:

• Watch the state updates in real-time
• See periodic AI analysis of application state
• Inspect logs through the dashboard

---

2. web_server.py - AI-Powered Web Server

What it demonstrates:

• Using .chat() to generate dynamic content
• Building an AI-native web API
• Tracking request metrics
• Inline AI generation

Run it:

reflexive --debug web_server.py

Try these URLs:

• http://localhost:8080/
• http://localhost:8080/story?topic=space+adventure
• http://localhost:8080/story?topic=mystery+detective

What's cool:

• Each story is generated by AI on-the-fly
• The AI has context about your app's state
• You can ask the AI questions via the CLI

Example interaction:

User (via CLI): "How many stories have been generated?"
AI: "Based on the state, 5 stories have been generated so far."

---

3. data_pipeline.py - Monitoring Example

What it demonstrates:

• Real-time monitoring of a data pipeline
• Using AI to analyze performance metrics
• Periodic health checks via .chat()
• Web UI for visualization

Run it:

reflexive --debug data_pipeline.py

Features:

• Simulates processing 100 data records
• Tracks success rate, errors, processing time
• AI analyzes performance every 20 records
• Full metrics available in web dashboard (http://localhost:3099)

What to watch for:

• How AI interprets error rates
• Performance recommendations from AI
• Real-time state updates

---

Code Patterns

Pattern 1: Inline AI Generation

import reflexive

r = reflexive.make_reflexive()

# Generate content dynamically
def handle_request(topic):
    content = r.chat(f'Write about: {topic}')
    return content

Use cases:

• Content generation APIs
• Dynamic responses
• AI-assisted processing

Pattern 2: State Monitoring

import reflexive

r = reflexive.make_reflexive({'web_ui': True})

# Track application metrics
r.set_state('requests.count', count)
r.set_state('errors.rate', error_rate)

# Ask AI for insights
analysis = r.chat('Should I be concerned about the error rate?')

Use cases:

• Application monitoring
• Performance analysis
• Anomaly detection

Pattern 3: Periodic Health Checks

import reflexive
import time

r = reflexive.make_reflexive()

while processing:
    process_batch()

    if iteration % 10 == 0:
        health = r.chat('Analyze system health')
        if 'warning' in health.lower():
            alert_team(health)

Use cases:

• Background job monitoring
• ETL pipeline health
• System diagnostics

Tips

1. State Management

Keep state small and informative:

# Good: Metrics and counters
r.set_state('users.active', 42)
r.set_state('cache.hit_rate', 0.95)

# Bad: Large objects
r.set_state('all_users', huge_list)  # Don't do this!

2. Logging Strategy

Use structured logging:

r.log('info', f'Processing batch {id}: {count} records')
r.log('warn', f'Slow query detected: {query_time}ms')
r.log('error', f'Failed to connect: {error}')

3. Chat Questions

Be specific:

# Vague
r.chat('How are things?')

# Specific
r.chat('What is the error rate over the last 10 minutes?')
r.chat('Is memory usage trending up?')

4. Performance

Cache AI responses when appropriate:

# Cache static content
@lru_cache(maxsize=100)
def get_ai_description(category):
    return r.chat(f'Describe: {category}')

Next Steps

1. Try the examples with reflexive --debug
2. Modify them to fit your use case
3. Build your own AI-native app
4. Read the docs at ../README.md

Questions?

• 📚 Read the full API docs: ../README.md
• 🏗️ Understand the design: ../DESIGN.md
• 🆚 Compare with Node.js: ../COMPARISON.md
• 💬 Ask the AI: r.chat('How do I...')