Metered tools create a specific low-grade dysfunction: nobody knows the number. "Do we have credits for this?" becomes a Slack question aimed at whoever owns the account; heavy research days get planned on vibes; and the choice between running the deep analysis now versus next sprint gets made without the one input that should decide it.

The number is one free call away. Making it ambient — checked at sprint planning, checked before heavy days, checked by anyone without asking anyone — is less a workflow than a habit. But it's the habit that separates teams who use their research budget from teams who fear it.

get_credit_status returns the live state — wallet balance, plan tier, today's remaining free quota — to whoever asks, free, in any connected assistant. That's the whole mechanism; the value is in where the habit installs it:

Sprint planning ("what's the budget line this cycle?"), the morning of a heavy research day ("size today's ambitions"), mid-sprint pulse checks ("are we pacing?"), and the end-of-cycle read that turns usage into next cycle's plan. The Slack question to the account owner dies; in its place, a number everyone can see and plan against — which is what a budget is supposed to be.