Language models learned to read. EDK-1 learns the ground. Five things a machine must know to act on Earth, each building on those before it.
Grounding anything on Earth
A name, a sentence, a photograph or a coordinate resolves to one exact place, with everything known about it attached. Position is how a model stops guessing where and starts knowing.
Discover resolves a question to places. Inspect resolves a point.
POST /v1/discoverReading the state of a place
Terrain, risk, access and activity, fused into a single picture of what holds at a place. Condition turns ground truth into something a machine can decide from.
Inspect reads every known condition at a point. Discover finds the areas that meet them.
GET /v1/inspectUnderstanding how the world moves
Flows, rhythms and change over time: what a place is doing now, and where it is heading. Motion gives a model a sense of before and after.
Discover results carry when each thing was seen, so change can be read across them.
POST /v1/discoverWatching for change, everywhere
A standing watch over a place: the moment something new appears, it is noticed, described and sent on. Stream turns a model that answers into one that keeps an eye out.
Monitor scans an area for new events of the artifacts you choose and sends each one to your webhook. Tier 4 and above.
POST /v1/monitorsFour calls reach all five: Discover, Inspect, Route and Monitor.