{
  "schema": "varexis-public-evidence-record/0.3",
  "generated_at_utc": "2026-10-11T17:52:53.561Z",
  "record_id": "VXI-EXT-MS-SK-ORCH-02-v1",
  "record_kind": "external_reference",
  "title": "Agent workflows: Parallel tasks or a dependent sequence",
  "system": "Microsoft Semantic Kernel",
  "evidence_family": "agent_workflow_dependencies",
  "index_status": "PUBLIC_SAFE_SOURCE_REVIEW",
  "evidence_state": "PRIMARY_DOCUMENTATION_REVIEW_ONLY",
  "canonical_question": "Should agents run in parallel when one task needs another agent's output?",
  "near_questions": [
    "When should parallel agent tasks become a sequential workflow?",
    "When should an agent workflow be sequential?",
    "How do agent task dependencies affect orchestration?",
    "Can independent agent tasks run concurrently?",
    "Which pattern passes one agent's output to the next?",
    "Does parallel orchestration guarantee lower cost?",
    "Does concurrent orchestration guarantee faster answers?",
    "Can a pipeline stage run before its input exists?",
    "Is a shared task the same as a dependent pipeline?",
    "What distinguishes concurrent and sequential agents?",
    "Can an experimental workflow be assumed production ready?",
    "What evidence is needed before changing agent execution order?"
  ],
  "supported_claims": [
    {
      "claim": "Microsoft's Semantic Kernel guide describes concurrent orchestration as independent processing of a shared task, while sequential orchestration passes each result to the next agent in order. Choose the pattern around the actual data dependency; a later task that needs an earlier result is not independent. These orchestration features are documented as experimental. The pattern description establishes neither application correctness nor a measured speed or cost benefit.",
      "reliance": "DOCUMENTED_CAPABILITY_ONLY"
    }
  ],
  "not_demonstrated_claims": [
    "application performance or reliability from a pattern name",
    "independence of tasks that consume earlier outputs",
    "production maturity of experimental orchestration"
  ],
  "public_summary": "Microsoft's Semantic Kernel guide describes concurrent orchestration as independent processing of a shared task, while sequential orchestration passes each result to the next agent in order. Choose the pattern around the actual data dependency; a later task that needs an earlier result is not independent. These orchestration features are documented as experimental. The pattern description establishes neither application correctness nor a measured speed or cost benefit.",
  "related_record_ids": [],
  "source_review": {
    "source_type": "official_external_documentation",
    "source_title": "Microsoft: Semantic Kernel Agent Orchestration",
    "source_url": "https://learn.microsoft.com/en-us/semantic-kernel/frameworks/agent/agent-orchestration/",
    "checked_on": "2026-10-11",
    "candidate_record_id": "VXI-EXT-MS-SK-ORCH-02-v1",
    "candidate_record_sha256": "4a281d09b38ab15166b1e282e773b05407700a3c4fed5e48efbf69526fe62555",
    "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": "Microsoft's Semantic Kernel guide describes concurrent orchestration as independent processing of a shared task, while sequential orchestration passes each result to the next agent in order. Choose the pattern around the actual data dependency; a later task that needs an earlier result is not independent. These orchestration features are documented as experimental. The pattern description establishes neither application correctness nor a measured speed or cost benefit.",
    "required_buyer_context": "Input and output dependencies, installed SDK version, language support and measured behaviour in the intended workflow.",
    "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.",
  "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#agent-workflow-data-dependencies",
  "machine_readable_record": "/varexis-index/records/agent-workflow-data-dependencies.json"
}
