Managing Tilebox Datasets

Use this skill for operational and design work with Tilebox datasets: schema design, dataset creation/update, markdown documentation, collection management, datapoint queries with filtering, datapoint lookup, and generated types. Prefer the CLI for inspection and operations; consult docs and SDKs for ingestion.

Refresh CLI Metadata

Check exact installed flags and schemas before relying on memory:

tilebox agent-context dataset --output-schema

Relevant docs concepts:

• Datasets are strongly typed containers; every datapoint in a dataset follows the dataset schema.
• Dataset kinds add required fields automatically. Do not include required fields in the custom schema.
• Custom field descriptions and example values power automatic schema documentation.
• Existing fields cannot be removed or changed after data has been ingested. New fields can be added because fields are optional.
• Empty datasets are the exception: if all collections are empty, the schema can be freely edited.