Canonical Definitions

The atomic, immutable event layer from which the GASP revenue, retention and cost metrics derive. If a commercial event is not in the CEL, it does not exist for metric purposes. Operational metrics read the other datasets in the ontology.

The tagging layer that classifies CEL events by lifecycle stage, expansion type and cost function. ATL tags let each dual-lens metric compute its Operating form and its Market form from the same events.

The JSON mapping file that connects GASP's canonical vocabulary to your warehouse: a table for each GASP dataset you have, a column (`schema.table.column`) for each field and optional owners for field data and metric definitions. The adapter is the only file a consuming organization writes.

The combination of the GASP standard (metrics, relationships, ontology), your adapter (warehouse mapping) and an MCP server (agent access layer). Together they form a read-only semantic model of how a SaaS business operates. Agents query the twin instead of the source systems.

The structured representation of GASP's entities, concepts, fields, datasets, source categories, temporal semantics and validation rules. The ontology sits below the metric layer and provides the canonical vocabulary that adapters, queries and agents reference.

A classification that describes what kind of signal a metric provides. Every GASP metric has exactly one primary signal type; some carry a secondary.

GASP's recommended data sensitivity classification for each metric. Aligns with ISO 27001 and OSSA `data_classification` conventions.

GASP's recommended filtering pattern for each metric. Describes whether the metric is naturally filterable by user context.

Optional CEL field that captures the intent behind a revenue movement. Does not alter metric calculations. Enables downstream analysis of revenue quality (e.g. Adjusted NRR).

The customer classification by size, vertical or fit (SMB, Mid-Market, Enterprise or the organisation's own scheme), recorded in the `segment` field. Revenue, contract and amendment events carry the segment as of the event; the customer and subscription snapshots carry it as of the snapshot. Segment is a dimension for slicing metrics rather than an ATL tag. Pricing tier lives on the plan.

A decomposed representation of a metric's canonical formula, derived from formulas.json: explicit references to other metric nodes (with roles like numerator, denominator, addend) and the raw fields required. Enables agents to walk the dependency tree without parsing formula text.

The named team or person responsible for a metric definition in an organization's adapter. The adapter schema recommends an owner for every metric you report and the validator lists metrics without one. When a metric value is disputed, the owner resolves the dispute. GASP provides the canonical definition; the owner decides whether their organization follows it or deviates and documents any deviation.

GASP's structured description of a metric's time grain, aggregation method and reporting frequency. Each metric has a mapping (e.g. MRR = End-of-month snapshot, Daily sync).