Mock scenarios · No API key required
Train incident response reasoning,
not memorization.
You receive a realistic incident alert email. Write the plan you would execute as the on-call responder — containment, investigation, recovery, and evidence handling. A rule-based engine grades your response against a pre-paired rubric and explains what is strong, what is missing, and what is dangerous.
- Categories
- Phishing · Brute-force · Suspicious outbound
- Grading
- 8-category rubric · weighted
- Mode
- Deterministic · explainable
What this trains
Reasoning, not recall.
01
Controlled scenarios
Schema-guided, template-based generation — phishing, brute force, and suspicious outbound. Every scenario ships with its own rubric.
02
Rule-based grading
Deterministic keyword-group matching across 8 categories with explicit weights. No vague LLM opinions.
03
Explainable feedback
Per-category breakdown, detected dangerous actions, strengths, and missing items. You see exactly why you scored what you did.