{
  "format": "agent-article-v0",
  "article_id": "PM-GPT-CORE-05",
  "canonical_url": "/providers/pm-gpt-core-05",
  "page_type": "reference",
  "claim_ids": [],
  "evidence_ids": [
    "EVD-PM-GPT-CORE-05-URL-8E19A25035A2",
    "EVD-PM-GPT-CORE-05-URL-3A0020DF1048",
    "EVD-PM-GPT-CORE-05-URL-1A0CC2EB15CB",
    "EVD-PM-GPT-CORE-05-URL-AD7A3D152878",
    "PM-GPT-CORE-05-DOCUMENT-READBACK"
  ],
  "unknown_ids": [
    "PM-GPT-CORE-05-CURRENTNESS"
  ],
  "operational_object_ids": [],
  "provider": [
    "GPT"
  ],
  "models": [],
  "plans": [],
  "surfaces": [],
  "regions": [],
  "current_status": "UNKNOWN",
  "observation_status": "UNOBSERVED",
  "freshness": "UNKNOWN",
  "last_verified": null,
  "supersedes": [],
  "superseded_by": [],
  "provenance": [
    "DOC-083cde4caa627644-PM-GPT-CORE-05",
    "SRC-PM-GPT-CORE-05-URL-8E19A25035A2",
    "SRC-PM-GPT-CORE-05-URL-3A0020DF1048",
    "SRC-PM-GPT-CORE-05-URL-1A0CC2EB15CB",
    "SRC-PM-GPT-CORE-05-URL-AD7A3D152878"
  ],
  "article_sources": [
    {
      "id": "EVD-PM-GPT-CORE-05-URL-8E19A25035A2",
      "evidence_class": "RESEARCH",
      "title": "Source: OpenAI, Human-in-the-loop, [",
      "href": "https://openai.github.io/openai-agents-python/human_in_the_loop/",
      "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": "APPROVED_DOCUMENT_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-PM-GPT-CORE-05-URL-8E19A25035A2"
      ],
      "supporting_locators": [
        "Bounded mapping",
        "Source: OpenAI, Human-in-the-loop, [https://openai.github.io/openai-agents-python/human_in_the_loop/"
      ],
      "scope": "Bounded mapping",
      "provenance_status": "APPROVED_DOCUMENT_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-PM-GPT-CORE-05-URL-3A0020DF1048",
      "evidence_class": "RESEARCH",
      "title": "Source reuse: EXT-C04-015, OpenAI, Models, [",
      "href": "https://openai.github.io/openai-agents-python/models/",
      "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": "APPROVED_DOCUMENT_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-PM-GPT-CORE-05-URL-3A0020DF1048"
      ],
      "supporting_locators": [
        "Bounded mapping",
        "Source reuse: EXT-C04-015, OpenAI, Models, [https://openai.github.io/openai-agents-python/models/"
      ],
      "scope": "Bounded mapping",
      "provenance_status": "APPROVED_DOCUMENT_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-PM-GPT-CORE-05-URL-1A0CC2EB15CB",
      "evidence_class": "RESEARCH",
      "title": "通信locator: [",
      "href": "https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view",
      "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": "APPROVED_DOCUMENT_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-PM-GPT-CORE-05-URL-1A0CC2EB15CB"
      ],
      "supporting_locators": [
        "Auxiliary disposition",
        "通信locator: [https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view"
      ],
      "scope": "Auxiliary disposition",
      "provenance_status": "APPROVED_DOCUMENT_REFERENCE_PROJECTED"
    },
    {
      "id": "EVD-PM-GPT-CORE-05-URL-AD7A3D152878",
      "evidence_class": "RESEARCH",
      "title": "通信locator: [https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view](https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view) (modified 2026-09-08T11:",
      "href": "https://drive.google.com/file/d/17B5Gyqnsy8IwVMLlUsTEnfgoTi_5rkb7/view",
      "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": "APPROVED_DOCUMENT_REFERENCE",
      "source_or_experiment_ids": [
        "SRC-PM-GPT-CORE-05-URL-AD7A3D152878"
      ],
      "supporting_locators": [
        "Auxiliary disposition",
        "通信locator: [https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view](https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view) (modified 2026-09-08T11:40:03.525Z); [https://drive.google.com/file/d/17B5Gyqnsy8IwVMLlUsTEnfgoTi_5rkb7/view"
      ],
      "scope": "Auxiliary disposition",
      "provenance_status": "APPROVED_DOCUMENT_REFERENCE_PROJECTED"
    },
    {
      "id": "PM-GPT-CORE-05-DOCUMENT-READBACK",
      "evidence_class": "IMPL",
      "title": "原資料との収録照合",
      "supports": [
        "資料単位の本文ハッシュ、段落数、表数を照合しています。"
      ],
      "limit": "収録照合は主張の真偽や外部根拠の最新性を追加保証しません。",
      "origin_class": "DOCUMENT_INGESTION_READBACK",
      "source_or_experiment_ids": [
        "DOC-083cde4caa627644-PM-GPT-CORE-05"
      ],
      "supporting_locators": [
        "18WPldGemks2px0ukhKYINeBMrBwAmlAi",
        "083cde4caa627644036a2225f1f6ff29b8015bd902614fe1f7a00a997b4eb59c",
        "provider_mappings/PM-GPT-CORE-05.md",
        "80fde87fa2bf6035963dcff34bfc2a69ae12b834ba9ac11cb8092d3420fdd889"
      ],
      "scope": "公開指定を受けた資料単位の全文・表・出典。",
      "provenance_status": "SOURCE_IDENTITY_VERIFIED"
    }
  ],
  "research_sections": [
    {
      "heading": "資料の識別情報",
      "text": "Source SHA-256 52ec16c371abc0775bb2ce0a3f38dda6c18089f22059e3790134b6e31867f5a0"
    },
    {
      "heading": "Product Mapping — GPT / CORE-05",
      "text": "PRODUCT_MAPPING_ID: PM-GPT-CORE-05-v0.1\n\nAS_OF: 2026-09-09\n\nSTATUS: PUBLIC_MAPPING_BOUNDED_COMPLETE / LIVE_AND_ORG_RESIDUALS\n\nLAYER: PRODUCT_MAPPING\n\nCORE: packets/CORE-05.md — unchanged"
    },
    {
      "heading": "Bounded mapping",
      "text": "| CLAIM | CORE CONTRACT | DOCUMENTED SURFACE | LIMIT |\n| --- | --- | --- | --- |\n| PM-C05-GPT-001 | CLM-C05-004/011 approval | Agents SDK: needs_approval、interruptions、call ID、arguments、RunState approve/reject/resume | call identityはcanonical operation digestと同じではない。trusted rendering・理解率・業務認可は別検証 |\n| PM-C05-GPT-002 | CLM-C05-004 individual review | programmatic approvalとrun内sticky approvalが可能。sticky decisionはstate保存後も残る | 毎回人が審査する保証ではない。再開時のtarget/arguments/policy差をapplicationで確認する設計は本研究の分析 |\n| PM-C05-GPT-003 | CLM-C05-006/007 retry | Models docsのRunner-managed retries / Safety boundaries。通常retry判断だけではreplay保護を回避できない | model request再送制御は任意toolのdedup、外部receipt、atomicity、compensation、exactly-onceを保証しない |\n\nSource: OpenAI, Human-in-the-loop, [https://openai.github.io/openai-agents-python/human_in_the_loop/](https://openai.github.io/openai-agents-python/human_in_the_loop/) ; accessed 2026-09-08; sections Marking tools that need approval / How the approval flow works / Automatic approval decisions / Long-running approvals. New source EXT-C05-001.\n\nSource reuse: EXT-C04-015, OpenAI, Models, [https://openai.github.io/openai-agents-python/models/](https://openai.github.io/openai-agents-python/models/) ; accessed 2026-09-08; Runner-managed retries / Safety boundaries, response_started / replay_safety / stateful_request / approve_unsafe_replay.\n\n既存release anchor v0.22.0 / 4df9ecfは履歴として保持するが、上記rolling docsの全挙動が当該packageに存在するとは未確認。固定版code照合はRECOVERY_ACTIVE。現在版保証とは表記しない。"
    },
    {
      "heading": "Auxiliary disposition",
      "text": "- AUX-AUD-20260908T2020-01: scopeが現在のCORE-05開始点と一致。AUD-C05-010としてcoverage残差を採用。\n\n- AUX-Q-20260908T2020-01 → Q-PM-C05-GPT-001: approval mappingは部分採用。UI effectivenessとfixed-version implementationは未閉鎖。\n\n- AUX-Q-20260908T2020-02 → Q-PM-C05-GPT-002: model replay/tool idempotencyの分離を採用。\n\n- AUX-Q-20260908T2020-03 → Q-PM-C05-GPT-003: permission / trajectory / external final-stateのsurface別照合を次単位へ。\n\n- AUX-E-20260908T2039-01 → EXT-C05-001: 原典直接readback後に採用。\n\n- AUX-E-20260908T2039-02 → EXT-C04-015再利用: 重複sourceを発番せず、CORE-05 relationを追加。\n\n通信locator: [https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view](https://drive.google.com/file/d/1xd19BGqlA1fYuCkX5UQTbM3xklW6h0i9/view) (modified 2026-09-08T11:40:03.525Z); [https://drive.google.com/file/d/17B5Gyqnsy8IwVMLlUsTEnfgoTi_5rkb7/view](https://drive.google.com/file/d/17B5Gyqnsy8IwVMLlUsTEnfgoTi_5rkb7/view) (modified 2026-09-08T11:45:07.032Z)。\n\n補助通信は発見経路・採否記録であり、外部claimの引用終端ではない。通信自体は変更していない。"
    },
    {
      "heading": "Remaining coverage / next bounded unit",
      "text": "Q-PM-C05-GPT-003から、GPT Actions / hosted MCP / Agents SDKのpermission・audit/trajectory・final-state境界を原典照合する。dry-run parity、malicious content、approval deception、undo/compensationはCore test intentを保持し、製品での対応可能性は未確認。固定版実装・関連issueの照合も残る。"
    },
    {
      "heading": "Evidence origin / execution",
      "text": "Evidence origin / execution"
    },
    {
      "heading": "Actions / hosted MCP bounded delta — 2026-09-09",
      "text": "原典確認日: 2026-09-08。rolling仕様であり固定SDK版の実装証明ではない。\n\n| CLAIM | SOURCE | DOCUMENTED SURFACE | ANALYSIS / LIMIT |\n| --- | --- | --- | --- |\n| PM-C05-GPT-004 | EXT-C05-002 | Actionsのx-openai-isConsequential=trueは毎回確認しAlways allowを提示しない。省略時GET=false、その他=true | 本研究の分析: GETが無副作用である証明ではない。確認UIと業務認可・理解・最終commitを分離 |\n| PM-C05-GPT-005 | EXT-C05-003 | hosted MCPはallowed_toolsでimport対象を限定。既定で共有前に承認要求。mcp_approval_requestとmcp_approval_responseを対応づけ、mcp_callに出力を記録 | 本研究の分析: import filterは下流IAMの代替ではない。承認・call出力は外部最終状態の独立readbackではない |\n\nQ-PM-C05-GPT-003は公開Actions/MCP surfaceのみ部分解決。SDK tracing、固定版code parity、issue照合、実機適用は継続。新規実測なし。\n\n本研究の既存CORE-05 local fixtureとuser実測の帰属を変更していない。今回は新規試験・再実行・provider API・tenant・業務操作なし。公開資料readbackのみ。\n\nCore変更なし、Gemini遡及監査不要。Organization Overlayのapprover、許可tool、retention、SLA、数値閾値はUNKNOWN。\n\n本packet全体の完了判定および全体公開gateは開けない。"
    },
    {
      "heading": "Fixed-release parity and Action Evidence — 2026-09-09",
      "text": "固定対象: openai-agents-python v0.22.0 / release commit 4df9ecf / RunState schema 1.16。確認日は2026-09-09。\n\n1. CLAIM: PM-C05-GPT-006; SOURCE: EXT-C05-004; EXT-C03-014 reuse; FIXED-VERSION FINDING: v0.22.0のHITL文書とrun_state.pyは、tool-call interruption、call別approve/reject、run内sticky decision、serialized approvals/tool invocations/hosted MCP approvals、canonical tool invocation identity、hosted MCPのserver label分離を実装・文書化; LIMIT / DISPOSITION: 固定版parityを確認。call identity／schema persistenceは本研究のcanonical operation digest、承認者理解、business authorization、外部commit証明ではない\n\n2. CLAIM: PM-C05-GPT-007; SOURCE: EXT-C05-005; EXT-C03-014 reuse; FIXED-VERSION FINDING: v0.22.0 tracing文書はrunner/model/tool/guardrail/handoff spans、trace/group/parent IDs、非同期batch exportと明示flush、sensitive-data capture control、custom processorsを記述。OpenAI tracingはZDR組織では利用不可; LIMIT / DISPOSITION: trajectory evidenceの固定版surfaceまで確認。trace/spanは外部systemのreceipt/readback/commit、改ざん不能な監査保管、tenant retentionを証明しない。入力・出力取得はprivacy riskにもなる\n\n固定版照合によりQ-PM-C05-GPT-001のcode parityとQ-PM-C05-GPT-003のSDK tracing残差は閉じた。Q-PM-C05-GPT-002の外部tool idempotency、dedup retention、receipt、compensation、exactly-onceと、全問いのlive tenant／human effectiveness／organization値は、公開SDK証拠では閉じない。これは欠落ではなくscope routingである。\n\nActions、hosted MCP、Agents SDKは別surfaceとして保持する。approval interruption、model replay protection、trajectory trace、external final-state evidenceを相互代用しない。dry-run parity、malicious external content、approval deception、undo/compensationについて、現行公開証拠から一様なGPT製品保証は確認できないためCore test intentと明示的UNKNOWNを維持する。"
    },
    {
      "heading": "Bounded decision",
      "text": "公開Product Mappingは境界付き完了。新規provider/API/tenant/tool/external-system実行はなく、既存local/user実測の帰属も変更していない。Core claimは変更していないためGemini遡及監査は不要。次はPM-GPT-CORE-06へ進み、CORE-05のlive/Organization residualsは同packetへ持ち込まない。"
    }
  ],
  "human_url": "/providers/pm-gpt-core-05",
  "markdown_url": "/agent/article.md?path=%2Fproviders%2Fpm-gpt-core-05",
  "json_url": "/agent/article.json?path=%2Fproviders%2Fpm-gpt-core-05"
}