Creates an exporter that writes one structured JSON line per span to standard output. This is useful for cloud-native environments where structured logs are collected from stdout by a log aggregator (e.g., Fluentd, Vector, CloudWatch Logs).
Details
Each line is a self-contained JSON object containing span-level fields:
trace_id, span_id, parent_id, name, type, start_time,
end_time, status, duration_secs, plus any token/cost data and
attributes.
Examples
exp <- json_stdout_exporter()
tr <- Trace$new("demo")
tr$start()
s <- Span$new("step1", type = "tool")
s$start()
s$end()
tr$add_span(s)
tr$end()
export_trace(exp, tr)
#> {"trace_id":"9479e4f230c8c16c2e9720e08f99a419","span_id":"a7134dbb6c3332de","parent_id":null,"name":"step1","type":"tool","start_time":"2026-03-10T16:23:40.460Z","end_time":"2026-03-10T16:23:40.461Z","status":"ok","duration_secs":0.0004,"input_tokens":0,"output_tokens":0,"model":null,"error":null}