Analysis
In this section
| Page | What it covers | Code | Cowork | Desktop chat | VS Code |
|---|---|---|---|---|---|
| Trace dependencies | Find what is around an element, how two elements connect, and what depends on an element along one relationship, before you change or retire it. | Yes | Partly | With you | With you |
| Find out what changed, when and by whom | See what was created or changed in a period and who was active, and confirm it against a baseline comparison. | Yes | Yes | With you | With you |
| Compare declared and observed usage | Compare what a modeling technology declares (element and connector stereotypes, base types, tags, permitted values) with what the repository actually uses, and report the gaps as findings. | Yes | No | No | With you |
| Trace the impact of a change in Prolaborate | Find what is connected to an element, one hop at a time, and report what is affected together with exactly how far the search went. | Yes | Yes | Yes | Yes |
| Assess the impact of retiring a system | Before you retire or replace a system, find everything that depends on it, who owns what is affected and how critical it is, and where it is drawn, as an impact report you can take to the decision. | Yes | Yes | With you | With you |
| Check traceability coverage | Find out what fraction of your requirements, or any other set of elements, is realized in the model, and list the gaps, with the coverage rule and the count stated so the figure can be trusted. | Yes | Yes | With you | With you |
| Compare the as-is and to-be architecture | See exactly what a target architecture adds, removes and changes compared with the current one, and what the removals affect, whichever way your team models as-is and to-be. | Yes | Yes | With you | With you |
Where it runs: Yes = runs fully · With you = runs, and you do the steps in EA’s window · Partly = one part needs a capability the client lacks · No = not available · TBC = to be confirmed
Analysis is answering questions about the architecture from the model itself: what depends on what, how a requirement reaches the systems that realize it, what a retirement would affect, how much of the model is traced, what a target architecture changes, what changed and who changed it, and where the modeling language and the model disagree. The answers come from operations that read the live repository and are deterministic: the same question on the same model state gives the same answer. What varies is the question the AI asks for you, so every scenario here ends with how to check that the answer is complete.
Questions you can answer, and where to start
| Question | Start with | How you know the answer is complete |
|---|---|---|
| What depends on this application before I change or retire it? | Assess the impact of retiring a system | Neither trace was truncated, the connector types followed are named, and a one-hop count equals the element's dependency connectors |
| How does this requirement reach the systems that realize it? | Trace dependencies | Each path can be walked connector by connector, and the direction followed was the right one |
| Which requirements have no realizing element, and what fraction is covered? | Check traceability coverage | Covered plus uncovered equals the denominator, the rule is stated, and a sample cross-check agrees with the query |
| What does the target architecture add, remove and change, and what do the removals affect? | Compare the as-is and to-be architecture | Every element is accounted for as added, removed, changed, matched or unmatched, and the method is stated |
| What was created or changed in a period, and who was active? | Find out what changed, when and by whom, with Protect a change with a baseline for a before-and-after comparison | The item count matches a direct count for the same window; deletions show only in a baseline comparison |
| Where do my modeling language and my model disagree? | Compare declared and observed usage | Every declared and every observed stereotype in the scope was compared |
| What is connected to this element, seen from Prolaborate? | Trace the impact of a change in Prolaborate | The report states the depth reached and why the search stopped |
How to read these scenarios: How to read a scenario. Examples use Westbrook Bank as an illustration. Substitute your own model, technology and paths. Run the start-of-session check (Start a session safely) before any write.