ts-coach-model

To generate AI context and synonyms for a ThoughtSpot model, run this Skill against the model. It analyzes dependent Liveboards, Answers, and mined prose to propose consistent coaching assets without overwriting existing work.

ThoughtSpot model coaching optimizes Spotter performance by extracting evidence to propose AI context, synonyms, and reference questions. You need it when preparing new or updated models to ensure accurate natural language query responses.

You review proposed reference questions and business terms through a per-surface review workflow. The Skill generates an actionable review plan for Step 7, requiring explicit user confirmation before applying any changes to existing content.

Yes, you can generate coaching assets without losing existing configurations. The Skill enforces guardrails to avoid overwriting existing content and requires explicit user confirmation before applying any proposed changes.

When data signals are incomplete during Spotter model coaching, the Skill gracefully degrades. It extracts available evidence from model descriptions and dependent objects to produce candidate content within the available constraints.

Yes, ts-coach-model supports staged generation for coaching surfaces. It produces paraphrase variants and allows per-surface review to systematically improve Spotter performance across AI context, synonyms, and data model instructions.