Every vehicle data vendor publishes a coverage claim, and most of them are make lists. A make list answers a question nobody has. Whether a vendor covers Honda is not useful; whether it has the fluids, intervals and priced parts for the exact Honda on the lift is the whole question.
So coverage here is a property of a configuration, and the response says so in two fields you can read at runtime.
{ "config_key": "2019_honda_cr_v_ex_l15be", "enrichment": { "status": "complete", "fill_rate": 92 } }Status, then fill rate
Status says whether the configuration has been through enrichment at all. Fill rate says how much of the expected field set came back for it. A complete configuration at a high fill rate is one you can quote from without checking. A lower one is not an error; it means some groups are thinner for that car, and a client that surfaces the number lets a service advisor decide rather than discover it mid-quote.
What to do with a vehicle that is not covered
GET /v1/configs lists the enriched configurations for a year, make and model, with the status and fill rate for each, so you can check before you promise anything. If nothing comes back, the dataset does not hold that car yet, and POST /v0/enrich schedules it. Enrichment typically completes in minutes rather than days, and polling costs nothing.
This is also why the coverage page on this site talks about configurations rather than badges. A number that counts makes is easy to publish and hard to use.