Standards and instrumentation
MCP carries W3C trace context in each request’s params._meta as unprefixed traceparent, tracestate, and baggage keys, per SEP-414 (Final). As of 2026-11-16, the official TypeScript, Python, and C# SDKs inject and extract it with OpenTelemetry, the PHP, Rust, and Ruby SDKs read and pass the keys without emitting spans themselves, and no support was found in the Go or Java SDK source.
- OpenTelemetry GenAI Semantic Conventions - Repository that now owns the GenAI spans, metrics, and events, moved out of the main semantic-conventions repo.
- OpenTelemetry MCP semantic conventions - Client and server span and metric conventions for MCP requests, naming spans after the method and target (such as
tools/call get-weather) and carrying trace context inparams._meta; status Development. - OpenTelemetry GenAI agent spans - Span conventions for creating an agent (
create_agent {gen_ai.agent.name}) and invoking one (invoke_agent), with their attributes; status Development. - OpenTelemetry Semantic Conventions - Parent repo for all OTel conventions, including the HTTP, RPC, and database spans an agent stack also emits.
- OpenTelemetry GenAI Observability SIG - Charter and scope of the OTel project group driving GenAI telemetry; the place to track where the conventions are heading.
- OpenTelemetry Python GenAI instrumentation - Upstream instrumentation packages for OpenAI, Anthropic, Google GenAI, LangChain, Bedrock, and several agent frameworks.
- OpenLLMetry - OpenTelemetry-based auto-instrumentation for Python LLM apps, exporting to any OTel backend.
- OpenLLMetry-JS - The same instrumentation for TypeScript and Node.js applications.
- OpenInference - OpenTelemetry-compatible tracing spec and instrumentation libraries for Python, JavaScript, Java, and Go from the Arize Phoenix team.
- OpenLIT - OTel-native auto-instrumentation SDK for LLM, vector-database, and GPU calls, with a self-hosted UI to read the resulting traces.