Understand the guidance behind your AI.
Test an intent, follow the MCP's routing gates, and inspect the exact guidance it returns.
Sign in with MicrosoftYour workspace for inspecting guidance, workflows, commands, performance evidence, and recorded calls.
From request to guidance
- Intent classification
- Jev evaluates authored questions to identify the task's domain and context labels.
- Canonical gates
- Always-on guidance, explicit workflows, and available integrations constrain selection.
- Relevance and depth
- When configured, Jev evaluates which eligible items apply and whether full guidance is needed.
- Final capsule
get_guidancereturns the selected guidance.bstack_resolveremains its compatibility alias.
bstack returns guidance. Your AI client applies it. Test intents and detailed traces are not stored in the dashboard's call logs.