[jeffhandley]

Draft pending the first release of
open-telemetry/semantic-conventions-genai.
Both conventions are Development stability and unreleased. Merge once they ship in a tagged release.

What this PR implements

Area	Convention	Upstream	Compensating change
gen-ai	gen_ai.request.reasoning.level	semantic-conventions-genai#258	Emit on chat spans from ChatOptions.Reasoning.Effort in OpenTelemetryChatClient (normalized lowercase token; exact provider wire string may differ, e.g. OpenAI xhigh).
gen-ai/agent	gen_ai.execute_tool.duration	#201 + #322	New histogram recorded in FunctionInvocationProcessor (gen_ai.tool.name, gen_ai.tool.type, error.type). Meter shared via GetService(typeof(Meter)) on the chat + realtime OTel clients.

Validation: build clean (net8.0/net9.0/net10.0, 0 warnings); 677 AI + 103 realtime tests pass. No public API surface change (constants/helpers internal); doc-comment version reference left at v1.41 (GenAI repo has no release to bump to yet).

Upstream scan tracking

Legend: 🔴 implemented here · ✅ already aligned · 🟡 watch/deferred · 🟢 not applicable (no client / docs-only)

Merged upstream changes (Unreleased) -- applicability to dotnet/extensions

Upstream PR	Area	Change	Applicability	Status
#258	gen-ai	gen_ai.request.reasoning.level	🔴	Implemented
#201	gen-ai/agent	execute_tool.duration + invoke_agent.duration	🔴	execute_tool.duration implemented; invoke_agent.duration N/A (no invoke_agent span)
#322	gen-ai/agent	align invoke_agent.internal/execute_tool attrs+metrics	🔴	Informs execute_tool.duration tag set (implemented)
#217	gen-ai	top_k double->int, split retrieval.top_k	✅	ChatOptions.TopK already int?
#257	gen-ai	limit system_instructions to text parts	✅	Emits Instructions as a single text part
#219	gen-ai	conversation.id no fallback UUIDs	✅	Uses real conversation id only
#142	gen-ai	document modality in message schema	🟡	Watch (additive message serialization)
#162	gen-ai	conversation.compacted + CompactionPart	🟡	Deferred (no compaction API)
#179	gen-ai	prompt.version / prompt.variable	🟡	Deferred (no prompt-template API)
#211	gen-ai	billed vs consumed token counts	🟡	Watch (clarification)
#214	gen-ai	provider.name cond-required on operation.duration	🟡	Watch (we always set provider.name)
#212	gen-ai	generalize provider.name description	🟢	Docs-only
#216	gen-ai	span covers duration incl. retries	🟢	Docs-only
#99	gen-ai	add moonshot_ai provider value	🟢	provider.name is passthrough
#126	gen-ai/agent	workflow.duration metric	🟢	No invoke_workflow span emitted here
#97	gen-ai/agent	plan operation	🟢	No plan/agent span
#107	gen-ai/agent	agent.name sampling-relevant	🟢	No agent spans
#242	gen-ai/agent	limit agent.id to stable ids	🟢	agent.id not emitted
#289	gen-ai/agent	remove provider.name from internal agent spans	🟢	No internal agent span
#321	gen-ai/agent	invoke_agent.duration in-proc clarification	🟢	No invoke_agent span
#140	gen-ai	memory operation spans/attrs	🟢	No memory instrumentation
#136	mcp	tool.call.arguments/result opt-in	🟢	No MCP instrumentation
#220	mcp	context propagation (SEP-414)	🟢	No MCP instrumentation
#330	tooling	complex-attr JSON serialization templates	🟢	Upstream authoring only

In-flight upstream changes (open PRs) -- applicability if merged

Filter: open PRs proposing convention changes (excludes currently-open pure dependency/CI/chore PRs: #112, #290, #328, #340; since the prior scan, chore #342 merged and #282 closed unmerged).

Upstream PR	Area	Change	Applicability	Status
#98	gen-ai/agent	model A2A handoff as execute_tool span	🟡	Watch (we own execute_tool)
#215	gen-ai	clarify client.operation.duration scope	🟡	Watch (we emit it)
#197	gen-ai	modality/cache/phase token-usage breakdowns	🟡	Watch (token usage attrs)
#96	gen-ai	token.cache / token.reasoning metric attrs	🟡	Watch (token usage)
#144	gen-ai	input-messages BlobPart optional + stripped_reason	🟡	Watch (message content)
#143	gen-ai	byte_size on multimodal content parts	🟡	Watch (message content)
#341	gen-ai/agent	rename workflow.duration -> invoke_workflow.duration	🟢	N/A (not emitted here)
#164	gen-ai	server.inter_token_latency metric	🟢	N/A (server-side)
#336, #350, #270, #267, #252, #250, #238, #291, #202, #165, #351, #325	gen-ai/agent	agent entity / identity / finish_reason / invocation / server / authorization / threat / content-size + agentic reference scenarios	🟢	N/A (M.E.AI doesn't own agent/invoke_agent spans)
#185, #184	gen-ai/eval	evaluation operation/span + response.id	🟢	N/A (no eval instrumentation)
#195, #188, #190, #262	gen-ai	a2a protocol, workflow node, context-selection event, guardrail/security span	🟢	N/A (not instrumented here)
#283, #324	anthropic/aws-bedrock	provider reference scenarios	🟢	N/A (other SDK repos)

Tracking state

Upstream-Repo: open-telemetry/semantic-conventions-genai
Upstream-Scan-Ref: e153ed94728993acd0ee6a958559032ca8b20afe  # since 377226a: only chore #342 (agents.md), no convention change
Upstream-Scan-Date: 2026-06-26T16:06:34Z
Upstream-Release: none            # Unreleased; Towncrier fragments under changelog.d/
Core-Semconv-Dependency: v1.42.0
DotnetExtensions-Implemented-Version: v1.41

[dotnet-comment-bot]

‼️ Found issues ‼️

Project	Coverage Type	Expected	Actual
Microsoft.Extensions.Diagnostics.Testing	Line	99	98.65 🔻
Microsoft.Extensions.Telemetry	Line	93	91.95 🔻
Microsoft.Extensions.AI	Line	89	88.56 🔻
Microsoft.Extensions.AI	Branch	89	88.65 🔻
Microsoft.Extensions.AI.OpenAI	Line	75	62.89 🔻
Microsoft.Extensions.AI.OpenAI	Branch	75	50.41 🔻
Microsoft.Extensions.DataIngestion.MarkItDown	Line	75	4.46 🔻
Microsoft.Extensions.DataIngestion.MarkItDown	Branch	75	0 🔻
Microsoft.Extensions.Diagnostics.ResourceMonitoring	Line	99	96.03 🔻
Microsoft.Extensions.Diagnostics.ResourceMonitoring	Branch	99	94.39 🔻
Microsoft.Extensions.Diagnostics.ResourceMonitoring.Kubernetes	Line	99	97.73 🔻
Microsoft.Extensions.ServiceDiscovery.Dns	Line	75	69.93 🔻
Microsoft.Extensions.ServiceDiscovery.Abstractions	Line	75	42.11 🔻
Microsoft.Extensions.ServiceDiscovery.Abstractions	Branch	75	42.86 🔻
Microsoft.Extensions.ServiceDiscovery	Line	75	67.96 🔻
Microsoft.Extensions.ServiceDiscovery	Branch	75	71.43 🔻
Microsoft.Extensions.ServiceDiscovery.Yarp	Line	75	73.85 🔻
Microsoft.Extensions.ServiceDiscovery.Yarp	Branch	75	70 🔻
Microsoft.Extensions.VectorData.Abstractions	Line	75	37.39 🔻
Microsoft.Extensions.VectorData.Abstractions	Branch	75	22.73 🔻

🎉 Good job! The coverage increased 🎉
Update MinCodeCoverage in the project files.

Project	Expected	Actual
Microsoft.Gen.BuildMetadata	97	100
Microsoft.Gen.MetadataExtractor	57	73
Microsoft.Gen.MetricsReports	67	69
Microsoft.Extensions.AI.Abstractions	82	85
Microsoft.Extensions.AI.Evaluation.NLP	0	78
Microsoft.Extensions.Caching.Hybrid	82	84
Microsoft.Extensions.DataIngestion	75	89
Microsoft.Extensions.DataIngestion.Markdig	75	90
Microsoft.Extensions.Http.Resilience	97	100

Full code coverage report: https://dev.azure.com/dnceng-public/public/_build/results?buildId=1482921&view=codecoverage-tab