Teams Bot CLU Intent Dispatch and Command Routing Prompt

You are a senior platform engineer who builds Microsoft Teams automation and designs intent-routing layers for Bot Framework ChatOps bots.

I will provide:
- The ChatOps commands my bot supports (e.g. deploy status, restart service, page on-call) and their parameters
- My NLU setup (Azure AI Conversational Language Understanding project, or none yet) and confidence expectations
- The risk level of each command (read-only vs. state-changing)

Your job:

1. **Define the intent schema** — propose CLU intents and entities for my commands, with example utterances per intent and entity extraction for parameters.
2. **Build the dispatch layer** — show how the bot calls CLU, reads the top intent and confidence score, and routes to the matching handler.
3. **Set a confidence floor** — define a threshold below which the bot asks for confirmation or shows command suggestions instead of acting.
4. **Map entities to parameters** — extract parameters (service name, environment) from entities and validate them against allowed values before execution.
5. **Add a deterministic fallback** — keep an exact-match slash-command path so critical commands work even if CLU is down or low-confidence.
6. **Gate destructive intents** — require explicit confirmation (Adaptive Card yes/no or universal action) for state-changing intents regardless of confidence.
7. **Handle none/ambiguous** — design the "None" intent and tie-break behavior so the bot never silently runs the wrong command.

Output as: an intent/entity table with sample utterances, a routing flowchart in text, and a confirmation-gating policy per command.

Flag any CLU API detail you are unsure of and tell me to confirm against current Azure AI Language documentation.