View Issue Details

IDProjectCategoryView StatusLast Update
0005539CaseTalk ModelerUser Interfacepublic2026-04-20 15:01
ReporterMarco Wobben Assigned ToMarco Wobben  
PrioritynormalSeverityfeatureReproducibilityhave not tried
Status resolvedResolutionfixed 
Target Version15.0Fixed in Version15.0 
Summary0005539: AI Coach uses Socratic questioning and supports prompt customization
Description

The AI Coach now distinguishes between contexts where asking a question raises more awareness than giving an answer, and contexts where a direct answer is more useful.

For open-ended moments — opening a project, saving, editing, or requesting a manual assessment — the coach asks a single focused question to help the modeler reflect on what they have not yet considered. For expression validation and well-formedness results, the coach gives a direct assessment because the modeler needs a concrete answer, not a prompt to reflect.

All coaching prompts can be previewed in Options > Coach. The prompts are written to a configuration file on first run so that advanced users can adjust the wording to fit their domain or organization.

The chat sidebar also includes two new starters: 'Interview (Fact Modeling)', which turns the AI into a Socratic interviewer that only asks questions, and 'Critical Reading', which analyzes a text for what is not said.

TagsNo tags attached.
CaseTalk EditionCorporate

Activities

There are no notes attached to this issue.

Issue History

Date Modified Username Field Change
2026-04-17 15:29 Marco Wobben New Issue
2026-04-17 15:50 Marco Wobben Target Version => 15.0
2026-04-17 15:50 Marco Wobben Description Updated
2026-04-20 15:01 Marco Wobben Assigned To => Marco Wobben
2026-04-20 15:01 Marco Wobben Status new => resolved
2026-04-20 15:01 Marco Wobben Resolution open => fixed
2026-04-20 15:01 Marco Wobben Fixed in Version => 15.0