Skip to content

RTerm v3.2.14

Choose a tag to compare

View all tags
@DrOlu DrOlu released this 24 Aug 23:56
· 81 commits to main since this release
v3.2.14
e85f1c7

OTel fixes + OpenLLMetry-style LLM tracing

Fixed

  • OTLP metrics push was always empty — the push driver kept pushing the boot-time empty registry while the Prometheus renderer rebuilt a fresh one. Both now share buildHostMetricsRegistry(); collectors receive real host metrics every 30s.

New

  • OpenLLMetry-style LLM tracing — every model call (chat, thinking audits like task_completion_guard/self_correction, failures) becomes an APM span grouped per agent run. Spans carry model, operation, duration, token usage, finish reason, error status. Visible via observability:apmSummary and the dashboard APM section with zero config. With OTEL_EXPORTER_OTLP_ENDPOINT set, spans also forward as OTLP/HTTP JSON with standard gen_ai.* attributes for Jaeger/Tempo/Datadog.

Includes all chat-console stability fixes from v3.2.12.

Verification

  • typecheck:all green; backend-unit / layout-ui / automation suites green
  • Live: 3 task types ran clean (0 removed answers, exactly 1 DONE each)
  • Live APM ledger after those runs: llm.chat / llm.thinking / llm.audit.task_completion_guard spans with real latency percentiles

npm

neuralos@3.2.14 · rterm-backend@3.2.14 · rterm-cli@3.2.14

Assets 12
Loading