Skip to main content

Demos Overview

These demos illustrate IntentusNet's core capabilities with realistic scenarios. Each demo includes input, routing decision, policy evaluation, execution trace, and replay verification.

Available Demos​

DemoCapability Demonstrated
Dangerous Target FilteringPolicy filtering protects CCTV while allowing other systems
Crash RecoveryMid-execution crash recovery with deterministic resume
Model Swap PreventionDeterministic routing prevents behavioral drift

Demo Format​

Each demo follows a consistent format:

  1. Scenario: Real-world situation
  2. Input Intent: The request being made
  3. Route Result: Which agent(s) were selected
  4. Policy Decision: What was allowed/denied
  5. Execution Trace: Step-by-step events
  6. Replay Note: How replay applies

Running the Demos​

All demos are runnable Python scripts:

# Install IntentusNet
pip install intentusnet

# Clone examples
git clone https://github.com/Balchandar/intentusnet
cd intentusnet/examples/demos

# Run a demo
python dangerous_target_filtering.py
python crash_recovery.py
python model_swap_prevention.py

Key Takeaways​

After running these demos, you'll understand:

  • Policy filtering protects critical systems without blocking entire intents
  • Crash recovery identifies failure points and enables safe resume
  • Deterministic routing ensures consistent agent selection regardless of external changes
  • Replay returns exact recorded outputs for debugging and auditing

Next Steps​