{
  "format": "agent-article-v0",
  "article_id": "EA-051",
  "canonical_url": "/tools/ea-051-ai-telemetry",
  "page_type": "reference",
  "claim_ids": [],
  "evidence_ids": [
    "EVD-EA-051-URL-6A0029787A22",
    "EVD-EA-051-URL-5E36EEBF9907",
    "EVD-EA-051-URL-E45DC2C99A6D",
    "EVD-EA-051-URL-B8771AA21367",
    "EVD-EA-051-URL-08867ABA5640",
    "EVD-EA-051-URL-7889FF67F45A",
    "EVD-EA-051-URL-9FE096EE4EF7",
    "EVD-EA-051-URL-6ACAC33977CC",
    "EVD-EA-051-URL-B669B429143C",
    "EVD-EA-051-URL-66CB031FFA27",
    "EVD-EA-051-URL-D66257F5E85E",
    "EVD-EA-051-URL-0F5EEBDD1974",
    "EVD-EA-051-URL-C10087515A45",
    "EA-051-R11-RESEARCH",
    "EA-051-R11-WEB-READBACK"
  ],
  "unknown_ids": [
    "EA-051-UNKNOWN-01",
    "EA-051-UNKNOWN-CURRENTNESS"
  ],
  "operational_object_ids": [],
  "provider": [],
  "models": [],
  "plans": [],
  "surfaces": [],
  "regions": [],
  "current_status": "UNKNOWN",
  "observation_status": "SPEC_ONLY",
  "freshness": "UNKNOWN",
  "last_verified": null,
  "supersedes": [],
  "superseded_by": [],
  "provenance": [
    "COMM-052-EA0051",
    "SRC-EA-051-URL-6A0029787A22",
    "SRC-EA-051-URL-5E36EEBF9907",
    "SRC-EA-051-URL-E45DC2C99A6D",
    "SRC-EA-051-URL-B8771AA21367",
    "SRC-EA-051-URL-08867ABA5640",
    "SRC-EA-051-URL-7889FF67F45A",
    "SRC-EA-051-URL-9FE096EE4EF7",
    "SRC-EA-051-URL-6ACAC33977CC",
    "SRC-EA-051-URL-B669B429143C",
    "SRC-EA-051-URL-66CB031FFA27",
    "SRC-EA-051-URL-D66257F5E85E",
    "SRC-EA-051-URL-0F5EEBDD1974",
    "SRC-EA-051-URL-C10087515A45",
    "WEBBODY-EA0051-R8-940A0AA1"
  ],
  "article_sources": [
    {
      "id": "EVD-EA-051-URL-6A0029787A22",
      "evidence_class": "RESEARCH",
      "title": "author/org: W3C",
      "href": "https://www.w3.org/TR/trace-context/",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-6A0029787A22"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://www.w3.org/TR/trace-context/"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-5E36EEBF9907",
      "evidence_class": "RESEARCH",
      "title": "author/org: OpenTelemetry",
      "href": "https://opentelemetry.io/docs/specs/semconv/",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-5E36EEBF9907"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://opentelemetry.io/docs/specs/semconv/"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-E45DC2C99A6D",
      "evidence_class": "RESEARCH",
      "title": "author/org: OpenTelemetry",
      "href": "https://github.com/open-telemetry/semantic-conventions-genai",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-E45DC2C99A6D"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://github.com/open-telemetry/semantic-conventions-genai"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-B8771AA21367",
      "evidence_class": "RESEARCH",
      "title": "version/revision: repository main; schema URL shown as",
      "href": "https://opentelemetry.io/schemas/gen-ai/1.42.0",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-B8771AA21367"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "version/revision: repository main; schema URL shown as https://opentelemetry.io/schemas/gen-ai/1.42.0"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-08867ABA5640",
      "evidence_class": "RESEARCH",
      "title": "author/org: OpenTelemetry",
      "href": "https://github.com/open-telemetry/semantic-conventions-genai/releases",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-08867ABA5640"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/releases"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-7889FF67F45A",
      "evidence_class": "RESEARCH",
      "title": "author/org: OpenTelemetry",
      "href": "https://github.com/open-telemetry/semantic-conventions-genai/blob/main/docs/gen-ai/gen-ai-spans.md",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-7889FF67F45A"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/blob/main/docs/gen-ai/gen-ai-spans.md"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-9FE096EE4EF7",
      "evidence_class": "RESEARCH",
      "title": "author/org: OpenTelemetry",
      "href": "https://github.com/open-telemetry/semantic-conventions-genai/blob/main/docs/gen-ai/gen-ai-agent-spans.md",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-9FE096EE4EF7"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/blob/main/docs/gen-ai/gen-ai-agent-spans.md"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-6ACAC33977CC",
      "evidence_class": "RESEARCH",
      "title": "author/org: Arize AI / OpenInference",
      "href": "https://arize-ai.github.io/openinference/spec/",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-6ACAC33977CC"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://arize-ai.github.io/openinference/spec/"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-B669B429143C",
      "evidence_class": "RESEARCH",
      "title": "author/org: Arize AI / OpenInference",
      "href": "https://arize-ai.github.io/openinference/spec/semantic_conventions.html",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-B669B429143C"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://arize-ai.github.io/openinference/spec/semantic_conventions.html"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-66CB031FFA27",
      "evidence_class": "RESEARCH",
      "title": "author/org: Arize AI / OpenInference",
      "href": "https://arize-ai.github.io/openinference/spec/configuration.html",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-66CB031FFA27"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://arize-ai.github.io/openinference/spec/configuration.html"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-D66257F5E85E",
      "evidence_class": "RESEARCH",
      "title": "author/org: OpenTelemetry semantic-conventions-genai contributors",
      "href": "https://github.com/open-telemetry/semantic-conventions-genai/issues/94",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-D66257F5E85E"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/issues/94"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-0F5EEBDD1974",
      "evidence_class": "RESEARCH",
      "title": "author/org: OpenTelemetry opentelemetry-python-genai contributors",
      "href": "https://github.com/open-telemetry/opentelemetry-python-genai/issues/98",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-0F5EEBDD1974"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://github.com/open-telemetry/opentelemetry-python-genai/issues/98"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-EA-051-URL-C10087515A45",
      "evidence_class": "RESEARCH",
      "title": "author/org: Arize-ai/openinference contributors",
      "href": "https://github.com/Arize-ai/openinference/issues/3097",
      "supports": [],
      "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。",
      "origin_class": "SOURCE_PACKET_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-EA-051-URL-C10087515A45"
      ],
      "supporting_locators": [
        "EVIDENCE TABLE",
        "canonical URL: https://github.com/Arize-ai/openinference/issues/3097"
      ],
      "scope": "EVIDENCE TABLE",
      "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED"
    },
    {
      "id": "EA-051-R11-RESEARCH",
      "evidence_class": "RESEARCH",
      "title": "EA-051 監査済みresearch / canonical revision binding",
      "supports": [
        "- Stable common substrate and AI semantic layer must be separated. W3C Trace Context provides a mature cross-service propagation mechanism; OpenTelemetry provides transport/data-model/collector ecosystem. AI-specific semantics are moving faster and should be version-gated independently.\n- OpenTelemetry GenAI semantic conventions were split from the core semantic-conventions repository in 2026. The dedicated repository exposes a GenAI schema URL but, at the inspected releases page, no formal GitHub release exists. Therefore future-book wording must not imply a settled GA-level AI telemetry standard.\n- Current OpenTelemetry GenAI material covers not only inference but retrieval, memory, tool execution, agent/workflow/planning operations. This is sufficient to define a future Core trace topology, while individual attribute stability still requires per-version confirmation.\n- OpenInference is not an alternative transport to OpenTelemetry; it is an AI-specific semantic convention layer designed to ride on OpenTelemetry/OTLP. It already provides a broad span-kind taxonomy and evaluation/annotation concepts useful for production AI operations.\n- OpenTelemetry GenAI and OpenInference are overlapping but not identical schemas. Treating them as wire-compatible synonyms would be unsafe. An update should maintain an explicit mapping table and record the originating schema/version on telemetry.\n- Instrumentation maturity differs by provider/language. The cited OpenTelemetry Python GenAI issue demonstrates that even within the same project some integrations may emit different signal sets. “Uses OTel” does not prove equivalent observability coverage.\n- Agent hierarchy remains an active design area. Issue #94 records a missing generic grouping primitive for iterations/tasks/skills/phases, which limits generic reconstruction of complex agent trajectories.\n- Sensitive prompt/tool/output capture is a governance control, not a harmless debugging default. OpenInference exposes explicit masking knobs; the update should require capture policy, redaction policy and retention policy before enabling content-rich traces."
      ],
      "limit": "canonical監査・revision bindingの存在を示す。外部ソースcurrentnessが本文またはcontrol-planeで確定していない場合は、currentnessをUNKNOWNのまま保持する。",
      "origin_class": "AUDITED_CANONICAL_RESEARCH_PACKET",
      "source_or_experiment_ids": [
        "COMM-052-EA0051"
      ],
      "supporting_locators": [
        "COMM-052-EA0051",
        "EA-051-R5-LEGACY-MIGRATION-20261008-R1",
        "content/staging/EA-051.md"
      ],
      "scope": "監査済みEA本文・canonical revision・既存Source Graph relationの範囲。欠損currentnessは推測で閉じない。",
      "provenance_status": "AUDIT_PASS_CANONICAL_BINDING"
    },
    {
      "id": "EA-051-R11-WEB-READBACK",
      "evidence_class": "IMPL",
      "title": "EA-051 Web body identity / canonical readback",
      "supports": [
        "canonical staging bodyのSHA-256一致とR9 canonical ledger readbackをProduction projectionの入力として検証する。"
      ],
      "limit": "body identity/readbackは本文claim自体の真偽や外部ソースの最新性を追加保証しない。",
      "checked_at": null,
      "origin_class": "IMPLEMENTATION_READBACK",
      "source_or_experiment_ids": [
        "WEBBODY-EA0051-R8-940A0AA1"
      ],
      "supporting_locators": [
        "WEBBODY-EA0051-R8-940A0AA1",
        "content/staging/EA-051.md",
        "EA-051-R5-LEGACY-MIGRATION-20261008-R1"
      ],
      "scope": "R9/R10 canonical readback後のProduction生成時body identity。",
      "provenance_status": "CANONICAL_READBACK_BODY_HASH_MATCH"
    }
  ],
  "research_sections": [
    {
      "heading": "WHY_THIS_MATTERS_TO_UPDATE",
      "text": "Enterprise AIの本番運用では、単一providerのログ画面だけではmodel call、retrieval、tool、agent orchestration、application/backendを一つの因果系列として追跡できない。更新版では「OpenTelemetryで全部標準化済み」と単純化せず、W3C Trace Context / OpenTelemetry core / GenAI semantic conventions / OpenInference等の役割と成熟度を分離し、移行期の互換性・privacy・schema driftまで扱う必要がある。"
    },
    {
      "heading": "SCOPE_ROUTING",
      "text": "Enterprise AI Core / interoperability / observability / implementation / governance / security-privacy / provider mapping input"
    },
    {
      "heading": "CLAIM_OR_RESEARCH_QUESTION",
      "text": "AI telemetryをprovider横断で構成する際、2026-09-14時点で何が安定した共通基盤で、何が発展途上のAI固有semantic layerか。OpenTelemetry GenAIとOpenInferenceはどこまで共通化でき、どこをschema/version gateとして扱うべきか。"
    },
    {
      "heading": "EVIDENCE TABLE",
      "text": "1. source class: standard\n   exact title: Trace Context\n   author/org: W3C\n   canonical URL: https://www.w3.org/TR/trace-context/\n   publication/update/access: W3C Recommendation 2020-02-06; accessed 2026-09-14\n   version/revision: Recommendation with editorial updates\n   locator: traceparent / tracestate definitions; distributed trace context propagation\n   evidence label: VERIFIED_SPEC\n   relation: support/context\n\n\n2. source class: official specification\n   exact title: OpenTelemetry semantic conventions 1.44.0\n   author/org: OpenTelemetry\n   canonical URL: https://opentelemetry.io/docs/specs/semconv/\n   publication/update/access: accessed 2026-09-14\n   version/revision: SemConv 1.44.0 page; GenAI section points to dedicated repository\n   locator: “Generative AI: Moved to the OpenTelemetry GenAI semantic conventions repository.”\n   evidence label: VERIFIED_SPEC / PRODUCT_VOLATILE\n   relation: support/context\n\n\n3. source class: author-maintained repository/spec\n   exact title: OpenTelemetry GenAI Semantic Conventions\n   author/org: OpenTelemetry\n   canonical URL: https://github.com/open-telemetry/semantic-conventions-genai\n   publication/update/access: accessed 2026-09-14\n   version/revision: repository main; schema URL shown as https://opentelemetry.io/schemas/gen-ai/1.42.0\n   locator: README; GenAI-specific spans, metrics, events, MCP/provider conventions\n   evidence label: VERIFIED_SPEC / OSS_PRACTICE / PRODUCT_VOLATILE\n   relation: support/context\n\n\n4. source class: author-maintained repository release state\n   exact title: Releases · open-telemetry/semantic-conventions-genai\n   author/org: OpenTelemetry\n   canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/releases\n   publication/update/access: accessed 2026-09-14\n   version/revision: no GitHub releases visible as of access date\n   locator: releases page — “There aren’t any releases here”\n   evidence label: OBSERVED / PRODUCT_VOLATILE\n   relation: limit/counter\n\n\n5. source class: official specification source\n   exact title: Semantic conventions for generative client AI spans\n   author/org: OpenTelemetry\n   canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/blob/main/docs/gen-ai/gen-ai-spans.md\n   publication/update/access: accessed 2026-09-14\n   version/revision: repository main\n   locator: page status “Development”; inference / embeddings / retrievals / fetch response / memory / execute tool spans\n   evidence label: VERIFIED_SPEC / PRODUCT_VOLATILE\n   relation: support/limit\n\n\n6. source class: official specification source\n   exact title: Semantic conventions for GenAI agent calls\n   author/org: OpenTelemetry\n   canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/blob/main/docs/gen-ai/gen-ai-agent-spans.md\n   publication/update/access: accessed 2026-09-14\n   version/revision: repository main\n   locator: create agent / invoke workflow / plan span definitions and requirement language\n   evidence label: VERIFIED_SPEC / PRODUCT_VOLATILE\n   relation: support/context\n\n\n7. source class: author-maintained specification\n   exact title: OpenInference Specification\n   author/org: Arize AI / OpenInference\n   canonical URL: https://arize-ai.github.io/openinference/spec/\n   publication/update/access: accessed 2026-09-14\n   version/revision: current docs\n   locator: OpenInference described as AI observability semantic conventions built on OpenTelemetry; LLM, agent, tool, retrieval, evaluation and privacy requirements\n   evidence label: VERIFIED_SPEC / OSS_PRACTICE\n   relation: support/context\n\n\n8. source class: author-maintained specification\n   exact title: Semantic Conventions | OpenInference\n   author/org: Arize AI / OpenInference\n   canonical URL: https://arize-ai.github.io/openinference/spec/semantic_conventions.html\n   publication/update/access: accessed 2026-09-14\n   version/revision: current docs\n   locator: span kinds LLM / EMBEDDING / CHAIN / RETRIEVER / RERANKER / TOOL / AGENT / GUARDRAIL / EVALUATOR / PROMPT; trace/session evaluations\n   evidence label: VERIFIED_SPEC / OSS_PRACTICE\n   relation: support\n\n\n9. source class: author-maintained specification\n   exact title: OpenInference Configuration\n   author/org: Arize AI / OpenInference\n   canonical URL: https://arize-ai.github.io/openinference/spec/configuration.html\n   publication/update/access: accessed 2026-09-14\n   version/revision: current docs\n   locator: OPENINFERENCE_HIDE_* controls and __REDACTED__ behavior; defaults documented as False for multiple capture-hiding switches\n   evidence label: VERIFIED_SPEC / OSS_PRACTICE\n   relation: support/limit/security\n\n\n10. source class: author issue / failure-gap report\n    exact title: Generic Grouping Attributes for Agentic Workflow Spans #94\n    author/org: OpenTelemetry semantic-conventions-genai contributors\n    canonical URL: https://github.com/open-telemetry/semantic-conventions-genai/issues/94\n    publication/update/access: opened 2026-04-25; accessed 2026-09-14\n    version/revision: issue #94, open at indexed source\n    locator: lack of standardized grouping for iterations/tasks/skills/phases in agentic workflows\n    evidence label: SOURCE_REPORT / OSS_PRACTICE / PRODUCT_VOLATILE\n    relation: counter/limit\n\n\n11. source class: author issue / implementation-gap report\n    exact title: Add metrics support to Anthropic & Groq GenAI instrumentations #98\n    author/org: OpenTelemetry opentelemetry-python-genai contributors\n    canonical URL: https://github.com/open-telemetry/opentelemetry-python-genai/issues/98\n    publication/update/access: opened 2026-05-14; accessed 2026-09-14\n    version/revision: issue #98\n    locator: OpenAI v2 emits spans+metrics while cited Anthropic/Groq instrumentations emit spans without metrics\n    evidence label: SOURCE_REPORT / OSS_PRACTICE / PRODUCT_VOLATILE\n    relation: counter/limit\n\n\n12. source class: author issue / schema-operability gap\n    exact title: Publish Semantic Conventions in a Machine-Readable Format #3097\n    author/org: Arize-ai/openinference contributors\n    canonical URL: https://github.com/Arize-ai/openinference/issues/3097\n    publication/update/access: opened 2026-05-13; accessed 2026-09-14\n    version/revision: issue #3097\n    locator: conventions defined in Markdown and manually re-declared across language packages; request for canonical machine-readable representation\n    evidence label: SOURCE_REPORT / OSS_PRACTICE / PRODUCT_VOLATILE\n    relation: counter/limit"
    },
    {
      "heading": "FINDINGS",
      "text": "- Stable common substrate and AI semantic layer must be separated. W3C Trace Context provides a mature cross-service propagation mechanism; OpenTelemetry provides transport/data-model/collector ecosystem. AI-specific semantics are moving faster and should be version-gated independently.\n- OpenTelemetry GenAI semantic conventions were split from the core semantic-conventions repository in 2026. The dedicated repository exposes a GenAI schema URL but, at the inspected releases page, no formal GitHub release exists. Therefore future-book wording must not imply a settled GA-level AI telemetry standard.\n- Current OpenTelemetry GenAI material covers not only inference but retrieval, memory, tool execution, agent/workflow/planning operations. This is sufficient to define a future Core trace topology, while individual attribute stability still requires per-version confirmation.\n- OpenInference is not an alternative transport to OpenTelemetry; it is an AI-specific semantic convention layer designed to ride on OpenTelemetry/OTLP. It already provides a broad span-kind taxonomy and evaluation/annotation concepts useful for production AI operations.\n- OpenTelemetry GenAI and OpenInference are overlapping but not identical schemas. Treating them as wire-compatible synonyms would be unsafe. An update should maintain an explicit mapping table and record the originating schema/version on telemetry.\n- Instrumentation maturity differs by provider/language. The cited OpenTelemetry Python GenAI issue demonstrates that even within the same project some integrations may emit different signal sets. “Uses OTel” does not prove equivalent observability coverage.\n- Agent hierarchy remains an active design area. Issue #94 records a missing generic grouping primitive for iterations/tasks/skills/phases, which limits generic reconstruction of complex agent trajectories.\n- Sensitive prompt/tool/output capture is a governance control, not a harmless debugging default. OpenInference exposes explicit masking knobs; the update should require capture policy, redaction policy and retention policy before enabling content-rich traces."
    },
    {
      "heading": "LIMITS / COUNTEREVIDENCE / FAILURE ENVELOPE",
      "text": "- Repository main pages are moving targets; exact attribute names/status can change without product deployment changes.\n- “No GitHub releases” is an as-of observation, not evidence that the project is unusable or unversioned in every distribution channel.\n- OpenInference author docs establish its own intended semantics, not universal industry adoption.\n- Issue reports establish concrete gaps in cited implementations/revisions, not defects in every language/provider/version.\n- Trace completeness can fail at async boundaries, browser/desktop automation boundaries, external SaaS calls or components that do not propagate context. A trace backend cannot reconstruct events never instrumented.\n- Capturing full model inputs/outputs can create PII, confidential-data, prompt-secret and retention exposure. Redaction after export may be too late.\n- Sampling can distort rare-failure analysis and token/cost aggregation unless sampling semantics are accounted for."
    },
    {
      "heading": "AVAILABILITY_GATES",
      "text": "- Language SDK support and auto-instrumentation coverage differ by runtime.\n- OpenTelemetry GenAI attribute/status/release availability must be rechecked at update publication time.\n- Provider SDK versions determine which token/cost/tool attributes are observable.\n- Collector/backend must preserve required attributes and W3C trace context across service boundaries.\n- Organization policy may prohibit payload-level prompt/output telemetry even where SDK capture is technically available."
    },
    {
      "heading": "IMPLEMENTATION_STATE",
      "text": "Status: implementable now with maturity caveats.\nRecommended architecture: W3C Trace Context + OpenTelemetry SDK/OTLP/Collector as substrate; choose one AI semantic profile per instrumentation boundary; persist schema/profile version; map alternate schemas at collector/backend rather than silently mixing attribute names. Instrument at least model, retrieval, tool, agent/workflow and application/backend boundaries. Add evaluation signal separately from raw execution telemetry."
    },
    {
      "heading": "BENCHMARK_OR_EVAL",
      "text": "No benchmark score proves telemetry completeness. Validation should instead use a synthetic trace conformance test: execute a known model→retrieval→tool→model workflow; verify parent/child structure, context propagation, errors, retries, token metadata, evaluation attachment, redaction and sampling behavior. Repeat after instrumentation/provider SDK upgrades."
    },
    {
      "heading": "SECURITY / PRIVACY / AUDIT / COST / OPERABILITY",
      "text": "Security/privacy: classify prompts, completions, tool arguments/results and retrieved documents before capture; default to metadata-only where sensitivity is unknown. Redact before export when possible.\nAudit: preserve trace_id, span_id, schema/profile version, instrumentation library version, model/provider identifiers and evaluation provenance.\nCost: token/cost attributes are useful but provider instrumentation coverage is uneven; billing records remain the financial source of truth unless independently reconciled.\nOperability: deploy schema-change detection and dashboards that tolerate missing optional attributes rather than assuming every provider emits identical fields."
    },
    {
      "heading": "SAFE_WORDING_FOR_FUTURE_BOOK",
      "text": "“OpenTelemetry and W3C Trace Context provide a strong interoperability substrate for AI telemetry, but AI-specific semantic conventions remain an evolving layer. As of 2026-09-14, OpenTelemetry’s GenAI conventions have moved to a dedicated repository and several agentic concepts remain under active development. OpenInference offers a complementary OpenTelemetry-based AI schema with broad model/tool/retrieval/agent/evaluation coverage. Enterprise deployments should pin and record schema versions, validate provider/language instrumentation coverage, and treat prompt/output capture as governed sensitive telemetry.”"
    },
    {
      "heading": "UPDATE_WORK_HANDOFF",
      "text": "1. At update time, recheck OpenTelemetry GenAI repository release/tag/schema status and per-section stability.\n2. Build a mapping appendix: OTel GenAI ↔ OpenInference for model, retrieval, tool, agent/workflow, memory, evaluation and cost/token fields.\n3. Add a provider/language coverage matrix based on concrete instrumentation releases, not project-level capability claims.\n4. Define a minimal enterprise trace profile and synthetic conformance TEST_PROFILE.\n5. Connect this packet directly to EA-052 AgentOps/AI-SRE and later provider mappings."
    },
    {
      "heading": "OPEN_UNKNOWN",
      "text": "first formal release/versioning policy of semantic-conventions-genai; convergence/divergence trajectory with OpenInference; provider-specific auto-instrumentation parity; stable representation for agent iteration/task grouping."
    },
    {
      "heading": "STOP_CONDITION",
      "text": "bounded initial packet complete. Reopen when dedicated OTel GenAI formal release/stability changes, mapping work begins, or a provider mapping requires exact attribute-level verification.\n\n---"
    },
    {
      "heading": "AUDITED_BOUNDARY",
      "text": "W3C Trace Context/OpenTelemetry core substrate must remain separate from evolving AI-specific semantic layers. OpenTelemetry semantic-conventions v1.44.0 moved GenAI content to the dedicated semantic-conventions-genai repository; GenAI conventions remain actively changing/development and require schema/version pinning. OpenTelemetry GenAI and OpenInference are overlapping but not interchangeable schemas. Preserve provider/language instrumentation gaps and all OPEN_UNKNOWN."
    }
  ],
  "human_url": "/tools/ea-051-ai-telemetry",
  "markdown_url": "/agent/article.md?path=%2Ftools%2Fea-051-ai-telemetry",
  "json_url": "/agent/article.json?path=%2Ftools%2Fea-051-ai-telemetry"
}