Classifying Devices

Devices imported from external sources (such as Redfish BMC data) may not have a recognized device-type slug. The classify command matches unclassified devices against the device-types library.

Auto-Classify

# Automatically classify all unclassified devices
cani alpha classify --auto

Auto-classification uses manufacturer and model information from the imported data to find the best match in the device-types library.

Interactive Classification

Without the --auto flag, each unclassified device is presented interactively with candidate matches ranked by similarity:

# Walk through each unclassified device
cani alpha classify

Typical Workflow

Classification is usually needed after importing from Redfish or other raw sources:

# 1. Import from Redfish
cani alpha import redfish --root ./redfish-roots.json

# 2. Classify the imported devices
cani alpha classify --auto

# 3. Review the results
cani alpha show device

See Device Type Data Model for details on the device-types library schema.

Strict Mode

By default, cani runs in strict mode (--strict=true). In strict mode, every device must have a resolved device-type slug before export. Unclassified devices will cause the export to fail.

This ensures inventory data is always complete and portable across providers.

Bypassing Strict Mode

To allow exports with unclassified devices (e.g. during exploratory imports), disable strict mode:

# Disable strict mode for a single command
cani alpha export nautobot --strict=false

# Or set it globally
cani --strict=false alpha export csm ...

Disabling strict mode is useful when iterating on imports from new sources where not all device types are mapped yet.