{
  "schema_version": 1,
  "builder_plan": {
    "scenario": "Your coding assistant makes changes that still need substantial repair before merging.",
    "change": "Try the technique on held-out issues from a repository with reliable tests.",
    "baseline": "The same coding agent, repository snapshot, tools, and token budget.",
    "steps": [
      "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.",
      "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.",
      "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.",
      "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test."
    ],
    "measure": "Correct patches, regression failures, human repair time, and total cost per accepted change.",
    "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation."
  },
  "builder_review": null,
  "arxiv_id": "2610.18507",
  "title": "The Last Mile of Coding Agents for Long-Document Tasks: An Empirical Study on Clinical Trial Protocol Digitization",
  "summary": "The technique replaces a generic coding-agent harness (and reliance on frontier proprietary models) with a domain-specific harness that adds task skills and a verification loop for long-document-to-structured-output tasks. The cost is engineering effort to build domain skills and verifiers, plus extra verification passes that increase token consumption and latency, and the reported gains are measured only on clinical trial protocol digitization. After adoption, agents can still fail in the transcription and structuring stage, so structured outputs require downstream validation, and the skill assets may not transfer to other document domains without rework.",
  "rationale": "The core finding is directly actionable for a constrained team—a well-designed harness lets a locally served open-weight model substitute for expensive proprietary APIs—but the gains are demonstrated in one clinical domain, so the reader should validate the pattern on their own long-document extraction workload before production adoption.",
  "technique": "Domain-specific coding-agent harness with task skills and verification loop",
  "family": "agentes_de_codigo",
  "recommendation": "testar",
  "published": "2026-10-13",
  "as_of": "2026-10-18",
  "editorial_status": "indexed",
  "source_url": "https://arxiv.org/abs/2610.18507",
  "independent_implementations": 0,
  "total_implementations": 1,
  "stars": 10,
  "citations": null,
  "claims": [
    {
      "claim_id": "claim-01",
      "statement": "The technique replaces a generic coding-agent harness (and reliance on frontier proprietary models) with a domain-specific harness that adds task skills and a verification loop for long-document-to-structured-output tasks. The cost is engineering effort to build domain skills and verifiers, plus extra verification passes that increase token consumption and latency, and the reported gains are measured only on clinical trial protocol digitization. After adoption, agents can still fail in the transcription and structuring stage, so structured outputs require downstream validation, and the skill assets may not transfer to other document domains without rework.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.18507",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "claim_id": "claim-02",
      "statement": "TrialScribe improves a locally served open-weight model by 6.2–59.4 points across eight metrics, enabling it to match or surpass proprietary models such as GPT-5.6-Sol on protocol digitization; evidence recall is usually achieved, but last-mile transcription and structuring remain the dominant failure mode.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.18507",
      "source_page": null,
      "source_excerpt": ""
    }
  ],
  "exposure_map": [
    {
      "dimension": "quality",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "compute",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "latency",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "operations",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "compatibility",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "security",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "data_and_training",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "reproducibility",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    }
  ],
  "risks": [],
  "minimum_test": [],
  "open_questions": [],
  "independent_tests": [],
  "lab_reproductions": [],
  "report_available": false,
  "validation_tier": "unknown",
  "evidence_tier": "unknown",
  "infrastructure_basis": "unknown",
  "equations": [],
  "equations_status": "not_fetched",
  "core_kind": "",
  "equations_fetched_at": "",
  "equations_source_url": "",
  "equations_source_sha256": "",
  "exposures_reviewed_at": "",
  "exposures_source_sha256": ""
}