Runtime and Communication

Job Processing and Runtime Events

How long-running work is queued, executed, streamed, archived, and exposed back to API clients.

Pipelines are configured processing actions. A pipeline defines display metadata, optional parameter schema, and default resource requirements. A job is one execution of a pipeline against input data, usually an artifact bundle.

User-facing clients normally discover actions through GET /v1/spatial/bundles/{bundleId}/pipeline-actions, then create work with POST /v1/spatial/bundles/{bundleId}/jobs or POST /v1/jobs.

Job state is exposed through GET /v1/jobs and GET /v1/jobs/{jobId}. Runtime logs are exposed through GET /v1/jobs/{jobId}/events, GET /v1/jobs/{jobId}/events/stream, and GET /v1/jobs/{jobId}/archive.

When a pipeline succeeds, the worker uploads the output directory into a generated artifact bundle and the job row references it as outputBundleId.

Runtime and Communication

Core concepts for jobs, search, event delivery, and contextual communication flows.

Search

The document search model, indexing lifecycle, and query surfaces exposed by the API.