MiravoMiravo AI skill
Turn any agent into a Miravo expert. Describe your use case in plain language, and let the Miravo AI skill turn it into Miravo models and environments.
Start from the use case, not the schema. Describe the plant you need, the equipment it should include, and how it should behave. Let the Miravo AI skill draft the environment and any custom models behind it.
What you can build with it
New environments from scratch
Describe a packaging line, water plant, backup-power facility, or any other system. Let the skill turn that use case into a Miravo environment you can run, inspect, and refine.
Custom twin models
Ask for the equipment the built-in catalog does not already cover. Let the skill draft the .twin.yaml files instead of starting from a blank schema.
Faster iteration
Refine the draft in plain language. Ask for topology changes, added faults, older equipment, protocol changes, or more realistic behavior without rewriting the files by hand.
Prompt it in natural language
Start from the use case. Ask for the environment first. Ask for individual twin models only where the environment needs custom equipment or explicit overrides.
Use the Miravo AI skill to create a new environment from scratch for a food and beverage packaging line.
I need:
- infeed conveyors
- a rotary filler
- a capper
- a labeler
- a palletizer
- one buffer tank
- realistic line behavior, jams, wear, and named faults
- ISA-95 enterprise/site structure
- MQTT and OPC UA outputs
Reuse built-in models where they fit.
Create custom twin models only where the catalog is missing equipment.
Draft the full environment and any required sibling twin models.Use the Miravo AI skill to adapt a packaging-line environment for a beverage canning plant.
Keep the overall topology simple.
Add rinse, fill, seam, inspect, and pack stages.
One line should run close to nominal production.
One line should run older equipment with more lifecycle wear.
Reuse built-in models where possible.
Create local twin overrides only where behavior needs to change.Use the Miravo AI skill to create a new twin model for a rotary filler.
Include:
- start and stop behavior
- throughput and product count
- current draw tied to speed and load
- thermal lag
- jam detection
- lifecycle stages
- named faults
- engineering units
Make it suitable for a packaging-line environment.End-user workflow
- Describe the plant, machine, or use case in plain language.
- Let the skill decide which built-in environments and models already fit.
- Let the skill draft the environment entry and any missing twin models.
- Review the draft and ask for changes in natural language.
- Repeat until the environment matches the system you want to simulate.
Do not start from YAML unless you want to. Start from the real system. Describe the topology, assets, signals, faults, and protocols. Let the skill turn that into Miravo files.
What to include in your prompt
| Include this | Why it matters |
|---|---|
| Use case and operating context | Sets the system boundary and first model list |
| Topology and namespace shape | Shapes the environment structure and emitted paths |
| Equipment to reuse vs customize | Keeps the draft anchored to the built-in catalog |
| Signals, units, and expected behavior | Drives member design and generator choice |
| Commands, actions, and control surfaces | Determines methods and writable members |
| Lifecycle stages and named faults | Defines wear progression and failure behavior |
| Protocol outputs and runtime expectations | Shapes protocol scope and environment settings |
Related pages
Environments
Browse runnable built-in environments.
Custom content
Review environment folders and sibling overrides.
Templates
Review the .miravo.yaml environment schema.
Twin models
Review the .twin.yaml asset schema.
Was this page useful?
Last updated on