WHAT ASK IS
An LLM front-end on a fixed tooling layer: substantive answers are meant to come from structured tool results (summaries, samples, knowledge base), not free-form guessing. CoreFeedback, CoreProfile, CoreDatabase, and CoreDecisions remain the evidence; Ask argues from them. In CoreReports, Notebook mode keeps catalogue writes off while you draft.
HOW IT WORKS
Mode follows where you are
Feedback Explorer, Snippets, Clusters, Decision Surface, Game Database, Notebook — each mode exposes the topic tools that belong there, so the model is nudged toward the right data.
Citations and readiness
Review samples when you need quotes; summaries respect data readiness — if NLP is not ready, the product steers you to process first instead of speculating.
Metered turns
Assisted chat turns spend credits — see Pricing for Ask rules. Manual-only workflows in the Cores stay free.
CREDITS AT A GLANCE
Representative metering — CreditsCatalog in-app is authoritative.
| Action | Credits |
|---|---|
Assistant chat turn (per message you send) | 1 cr |
Rates may adjust before launch; VAT at checkout where applicable.
WHERE ASK SHOWS UP
Surfaces across CoreFeedback, CoreProfile, CoreDatabase, CoreDecisions, and CoreReports — same assistant, narrower tools per mode.
USE IT IN THE PRODUCT
Ships inside workspace flows — waitlist for access.