Assessments Overview
An assessment is a run of a framework against an engagement's evidence. It's the core analytical act of the Diagnose phase: it produces grounded, cited answers to the framework's questions, scores them, and surfaces the gaps the engagement will plan around.
Who can access
Members of the engagement team.
How an assessment works
- You pick a framework and confirm evidence coverage in Diagnose.
- The assessment runs over the engagement's retrieved evidence.
- It returns an answer for each question, each with a confidence score and citations.
- You review and accept answers; gaps are reviewed in Diagnose.
Key properties
- Versioned — every run is a version. Re-run as evidence improves and compare.
- Grounded — answers are generated from retrieved evidence, not invented. See Citations & Trust.
- Honest about gaps — a question with no supporting evidence is shown as Not grounded and recorded as a gap, rather than answered confidently with nothing behind it.
The sub-topics
| Topic | Page |
|---|---|
| The mechanics of starting and watching a run | Running Assessments |
| Getting evidence into the engagement first | Evidence Collection |
| Conversational evidence capture | AI Interviews |
| Turning findings into gaps | Gap Analysis |
| Reading scores and maturity | Maturity Scoring |