semantic-conventions

An open governance specification for AI workloads, compatible with OpenTelemetry.

Turn existing OpenTelemetry telemetry into a live, lifecycle-managed typed property graph.

OTel-native typed primitives for LLM cost attribution and telemetry — published on PyPI.

OpenTelemetry GenAI semantic conventions for Go: drop-in HTTP middleware, gRPC interceptors, and client transports that emit spec-aligned inference + tool spans with multi-turn conversation correlation.

A hands-on lab where OpenTelemetry Weaver defines, generates from and continuously enforces the telemetry schema of a small distributed system whose signals are collected by a two-tier Bindplane OpenTelemetry Collector pipeline.

OTel Collector processor that normalizes vendor/framework LLM attributes to canonical gen_ai.* semantic conventions. One dashboard to rule them all.

pytest assertions that verify OpenTelemetry spans conform to the gen_ai semantic conventions

A portable record binding what an agent proposed, what a human decided, what executed, and what it cost. An OpenTelemetry semantic convention plus a thin emitter. The record, not the enforcer.

Validate OpenTelemetry GenAI spans against the official semantic conventions: 10 rules covering required attrs (provider/operation), token usage names + ranges, legacy attr-name detection, span kind, name format. Library + CLI.

A deterministic conformance validator for GenAI and agent telemetry. Checks real OTLP traces against OpenTelemetry's GenAI semantic conventions: required gen_ai.* attributes, token/cost observability, content-capture safety, and agent-trace structure. Graded, CI-gateable, no LLM.

Reference implementation of OpenTelemetry GenAI semantic conventions — wrappers for Anthropic, OpenAI, Vertex AI, and Bedrock that emit spec-compliant spans. Dashboards for Phoenix, Langfuse, and Cloud Trace.

Curated OTel GenAI conformance corpus, categorized OTLP/JSON fixtures (valid / dirty / shape) with declarative manifest. Drop-in regression target for any GenAI consumer: validators, exporters, rollups, dashboards.

OpenTelemetry semantic conventions and instrumentation for agent provenance, derivation lineage, and acceptance criteria evaluation. Fills the Microsoft AI stack observability gap.

OpenTelemetry instrumentation package for the Dust SDK. Automatically instruments agent interactions using OTel GenAI Semantic Conventions.

A vendor-neutral OpenTelemetry semantic convention for data staleness (data freshness), plus a Python SDK and Go Collector components.

Spring AI chat app that queries Grafana over MCP and exports its own telemetry to the official OpenTelemetry Collector. Notes on where Spring AI diverges from the GenAI semantic conventions.

OpenTelemetry semantic conventions for instrumented post-mortem badgers.

SZL OTLP collector, signed-span exporter and Helm package. Restored source owner; runtime and forwarding acceptance verified separately.