{
  "schema": "varexis-public-evidence-record/0.3",
  "generated_at_utc": "2026-10-22T23:52:06.542Z",
  "record_id": "VXI-EXT-PROCUREMENT-LANGGRAPH-INTERRUPT-RESUME-PREAPPROVAL-REEXECUTION-01-v1",
  "record_kind": "external_reference",
  "title": "LangGraph interrupts: Resumption repeats preceding node code",
  "system": "LangGraph",
  "evidence_family": "vendor_evidence_review",
  "index_status": "PUBLIC_SAFE_SOURCE_REVIEW",
  "evidence_state": "PRIMARY_DOCUMENTATION_REVIEW_ONLY",
  "canonical_question": "Does resuming a LangGraph approval interrupt avoid repeating actions earlier in the same node?",
  "near_questions": [
    "Will LangGraph rerun code before interrupt when an agent resumes?",
    "Can a LangGraph approval node repeat an earlier API write?",
    "Where should side effects sit around a LangGraph human approval?",
    "Does a LangGraph checkpoint prevent duplicate pre-approval actions?"
  ],
  "supported_claims": [
    {
      "claim": "LangGraph resumes an interrupted node from its beginning, so code before the interrupt runs again. The guide requires preceding side effects to be idempotent and recommends placing effects after the interrupt or in separate nodes where appropriate. A saved checkpoint does not make arbitrary earlier effects execute only once.",
      "reliance": "DOCUMENTED_CAPABILITY_ONLY"
    }
  ],
  "not_demonstrated_claims": [
    "Exactly-once external effects, valid approval binding or tested recovery of a buyer workflow",
    "A conclusion that every replay repeats an external effect despite application deduplication",
    "Buyer deployment testing, live runtime enforcement, certification or permission to act",
    "Expanded capabilities or licences for existing INDEX executables"
  ],
  "public_summary": "LangGraph resumes an interrupted node from its beginning, so code before the interrupt runs again. The guide requires preceding side effects to be idempotent and recommends placing effects after the interrupt or in separate nodes where appropriate. A saved checkpoint does not make arbitrary earlier effects execute only once.",
  "related_record_ids": [],
  "source_review": {
    "source_type": "official_external_documentation",
    "source_title": "LangGraph interrupts",
    "source_url": "https://docs.langchain.com/oss/python/langgraph/interrupts",
    "checked_on": "2026-10-22",
    "candidate_record_id": "VXI-EXT-PROCUREMENT-LANGGRAPH-INTERRUPT-RESUME-PREAPPROVAL-REEXECUTION-01-v1",
    "candidate_record_sha256": "a21c6347df5228efb992d061bae4a3948c790e6a91b08cb71d79c9a8158b5cd1",
    "method": "Primary documentation reviewed for the published bounded statement.",
    "deployment_tested_by_varexis": false,
    "revalidation": "The delivered snapshot preserves the review date shown. Recheck current primary documentation and deployment context before relying on present applicability; payment does not trigger fresh research."
  },
  "answer_policy": {
    "classification": "DOCUMENTED_WITH_DEPLOYMENT_BOUNDARY",
    "bounded_answer": "LangGraph resumes an interrupted node from its beginning, so code before the interrupt runs again. The guide requires preceding side effects to be idempotent and recommends placing effects after the interrupt or in separate nodes where appropriate. A saved checkpoint does not make arbitrary earlier effects execute only once.",
    "required_buyer_context": "Node code ordering, interrupt location and workflow version; Thread identity, checkpoint and resume history; External-operation identity and idempotency or reconciliation evidence. Buyer deployment not inspected.",
    "missing_evidence": "A broader claim needs evidence matching the intended use. Payment does not upgrade the evidence state."
  },
  "private_boundary": "Raw logs, thresholds, private paths, source code, protected mechanisms and reconstruction-level details are excluded from public and paid delivery.",
  "decision_profile": {
    "schema": "varexis-vendor-evidence-review/1",
    "review_status": "documentation_review_only",
    "checked_on": "2026-10-22",
    "claim_to_check": "Does resuming a LangGraph approval interrupt avoid repeating actions earlier in the same node?",
    "documented_statement": {
      "status": "documented",
      "statement": "LangGraph resumes an interrupted node from its beginning, so code before the interrupt runs again. The guide requires preceding side effects to be idempotent and recommends placing effects after the interrupt or in separate nodes where appropriate. A saved checkpoint does not make arbitrary earlier effects execute only once.",
      "source_ids": [
        "langgraph-interrupts"
      ]
    },
    "review_inference": {
      "status": "review_inference",
      "statement": "A review needs the effect location and external deduplication evidence, not just the existence of a checkpoint.",
      "source_ids": [
        "langgraph-interrupts"
      ]
    },
    "buyer_evidence_required": [
      "Node code ordering, interrupt location and workflow version",
      "Thread identity, checkpoint and resume history",
      "External-operation identity and idempotency or reconciliation evidence"
    ],
    "review_condition": {
      "status": "varexis_review_policy",
      "statement": "Refuse an exactly-once or approval-before-effect claim without reviewing node ordering and prior external effects."
    },
    "unknowns": "The buyer node code, idempotency behavior and external action history are unobserved.",
    "unknown_finding": {
      "status": "unknown",
      "statement": "The buyer node code, idempotency behavior and external action history are unobserved."
    },
    "evidence_sources": [
      {
        "source_id": "langgraph-interrupts",
        "url": "https://docs.langchain.com/oss/python/langgraph/interrupts",
        "final_url": "https://docs.langchain.com/oss/python/langgraph/interrupts",
        "title": "LangGraph interrupts",
        "checked_on": "2026-10-22",
        "captured_at_utc": "2026-10-22T23:48:43.303Z",
        "version_context": "Rolling LangGraph Python documentation; interrupt/resume semantics, not a pinned installed package.",
        "sha256": "29c003beb884c29da7a4dc9b5b4ca4b2a88eb3e52704f67f9383fe574971763c",
        "bytes": 1523807,
        "hash_basis": "Decoded HTTP response body bytes; identifies captured content, not truth or certification",
        "reviewed_sections": [
          "Resuming interrupts",
          "Side effects called before interrupt must be idempotent"
        ],
        "capture_status": 200,
        "rights_review": "Factual paraphrase and source metadata only; response body retained privately and not redistributed."
      }
    ],
    "observed_deployment": false,
    "commercial_reliance_boundary": "Documentation review only; the buyer must establish applicability and actual outcomes. INDEX does not inspect or control the deployment.",
    "paid_value": {
      "decision_context": "Prevent repeated writes or purchases when an agent resumes after approval.",
      "free": "Complete review, sources, practical questions and limitations; may be saved locally.",
      "single_request": "Existing server-saved version-bound route snapshot for later authenticated retrieval.",
      "evidence_receipt": "The same snapshot plus the existing unsigned delivery receipt.",
      "excluded": [
        "Fresh vendor research",
        "Continuous monitoring or change alerts",
        "Testing the buyer configuration",
        "Private evidence or vendor certification"
      ],
      "integrity": "Hashes identify content versions; they do not prove truth or provide a signature."
    },
    "revalidation": {
      "trigger": "A saved purchase does not recheck the source or buyer configuration.",
      "specific_watch": "Recheck the documented behavior, product version and buyer evidence before relying on a current outcome."
    },
    "buyer_relevance": "Prevent repeated writes or purchases when an agent resumes after approval.",
    "refusal_conditions": [
      "Refuse an exactly-once or approval-before-effect claim without reviewing node ordering and prior external effects."
    ]
  },
  "niche": {
    "niche_id": "agent-authority-action-evidence",
    "label": "Agent Authority & Action Evidence",
    "classification_id": "audit",
    "classification_label": "Audit, Provenance & Evidence Accountability",
    "topic_id": "audit",
    "buyer_intent": "Prevent repeated writes or purchases when an agent resumes after approval.",
    "comparison_role": "supplementary_discovery"
  },
  "paid_access_note": "The full published answer and boundaries are free. Paid access saves the selected route and source versions, with an unsigned receipt when purchased. Preview coverage before paying. Fresh research, bespoke assessment and protected evidence are excluded.",
  "service_route": {
    "request_guide": "/varexis-index/request.html",
    "request_workflow": "/varexis-index/request-workflow.json",
    "service_id": "single_request",
    "service_routes_url": "/varexis-index/service-routes.json",
    "payment_required": true,
    "availability_endpoint": "/api/index/capabilities",
    "operational_status": "AUTOMATIC_SERVICE_CAPABILITIES_GATED",
    "checkout_url": "/index-access"
  },
  "authority_boundary": "VAREXIS-INDEX informs evidence reliance. It does not provide advice, certification, deployment authority, security assurance, production approval, or guarantee.",
  "human_page": "/varexis-index/catalog.html#langgraph-interrupt-resume-preapproval-reexecution",
  "machine_readable_record": "/varexis-index/records/langgraph-interrupt-resume-preapproval-reexecution.json"
}
