OpenEnv Curriculum RL
Server online
⚡ AI Endpoint may be cold-starting. The inference endpoint scales to zero when idle to save costs. If the first episode shows "AI timeout — using fallback", wait ~2 minutes and run again. Live AI inference will work once the endpoint is warm.
Scenario
Agent Mode
Speed
System State
Select a scenario and click
Run Episode to start.
Agent Action Feed
Agent actions appear here
during incident analysis.
// System logs will stream here...
Reward Curve
—
Final Score / 1.0
0.00
Cumul. Reward
0
Steps Taken
Diagnosis—
Action Efficiency—
Investigation Quality—
Scenario
Speed
Defender vs Attacker
Live System State
Click Start Battle
to launch live simulation.
Battle Feed
Attacker and defender actions
stream here in real time.
// Combat logs will appear here...
Battle Score
0.00
Defender
0.00
Attacker
—
Episode Score / 1.0
Defender Advantage—
Attack Suppression—
Battle Rounds0
Curriculum Progress
Run episodes to see
the self-improvement curriculum.
Learning Curve & Score History
Level-Up Events
No level-ups yet. Run episodes to progress!
How Self-Improvement Works
1. Start at Level 1 with foundational scenarios.
2. Track score over a rolling window of five episodes.
3. Promote level automatically when thresholds are met.
4. Increase attacker pressure as defender capability improves.
5. Level 5 reflects expert readiness for disguised exfiltration scenarios.
Result
Episode Complete
—