Readiness
The Readiness page answers the most important question: “Can I ship this AI?”Purpose
Readiness is the decision surface. It provides:- A clear ship/no-ship signal
- Status on all 10 readiness questions
- Visibility into what’s failing and why
What You See
Ship Confidence Score
A single number (0-100) representing overall ship-readiness. This is derived from:- Pass rates across all 10 readiness questions
- Severity weighting of failures
- Coverage completeness
The 10 Readiness Questions
Each question shows:- Current status (Pass / Warn / Fail)
- Score (0-100)
- Number of scenarios tested
| # | Question | What It Checks |
|---|---|---|
| 1 | Intent | Does it do the right thing? |
| 2 | Grounding | Is it truthful & grounded? |
| 3 | Hallucination | Did it hallucinate? |
| 4 | Rules | Did it follow our rules? |
| 5 | Safety | Did it avoid harm? |
| 6 | Consistency | Is it consistent? |
| 7 | Quality | Is it good enough? |
| 8 | Robustness | Is it robust to manipulation? |
| 9 | Brand Safety | Is it brand-safe? |
| 10 | Schema | Is the output structurally valid? |
Failing Scenarios
When something fails, you see:- Which scenario failed
- The input that triggered the failure
- The actual vs expected output
- LLM judge reasoning (why it failed)
Feature Map
A list of AI features detected in your codebase, with:- Feature name and location
- Number of scenarios covering it
- Current pass rate
User Flows
Readiness is central to these flows:- First-Time Setup - See initial status after discovery
- Debugging a Failure - Understand what went wrong
- Expanding Coverage - Identify gaps to fill
- Quarterly Safety Review - Export for leadership