diff --git a/alg/平整度、平面度、轮廓测量v1.5/pipeline.py b/alg/平整度、平面度、轮廓测量v1.5/pipeline.py
index 0fcf220..b2ead31 100644
--- a/alg/平整度、平面度、轮廓测量v1.5/pipeline.py
+++ b/alg/平整度、平面度、轮廓测量v1.5/pipeline.py
@@ -322,6 +322,7 @@ class MeasurementPipeline:
         excel_output_path: str | None = None,
         result_template_path: str | Path | None = DEFAULT_RESULT_TEMPLATE,
         combined_context_pointcloud_path: str | Path | None = DEFAULT_COMBINED_CONTEXT_POINTCLOUD,
+        save_combined_visualization: bool = True,
         no_visualize: bool = False,
     ) -> dict[str, Any]:
         batch_start_time = time.perf_counter()
@@ -443,7 +444,7 @@ class MeasurementPipeline:
 
         combined_visualization_output = None
         combined_pointclouds = []
-        if combined_visualizations:
+        if save_combined_visualization and combined_visualizations:
             combined_visualization_output = self._contour_combined_save_path(save_path)
             self._save_contour_combined_pointcloud_stream(
                 combined_visualizations,
@@ -627,6 +628,7 @@ class MeasurementPipeline:
         template_sheet_name, headers, template_rows = self._read_measurement_template(template_path)
         task_col = self._find_header_index(headers, ["任务号", "任务编号", "task_id"])
         result_col = self._find_header_index(headers, ["测量结果"])
+        error_col = self._find_header_index(headers, ["误差结果"])
         conclusion_col = self._find_header_index(headers, ["测量结论", "结论", "状态"])
 
         result_by_task = {
@@ -648,6 +650,7 @@ class MeasurementPipeline:
                 continue
 
             output_row[result_col] = self._row_measurement_mean(result_row)
+            output_row[error_col] = result_row.get("max")
             output_row[conclusion_col] = self._row_measurement_conclusion(result_row)
             table.append(output_row)
             used_task_ids.add(task_id)
@@ -659,6 +662,7 @@ class MeasurementPipeline:
             output_row = [""] * len(headers)
             output_row[task_col] = task_id
             output_row[result_col] = self._row_measurement_mean(result_row)
+            output_row[error_col] = result_row.get("max")
             output_row[conclusion_col] = self._row_measurement_conclusion(result_row)
             table.append(output_row)
 
@@ -729,7 +733,9 @@ class MeasurementPipeline:
 
     def _row_measurement_conclusion(self, row: dict[str, Any]) -> str:
         status = str(row.get("status", "")).strip()
-        if status and status != "成功":
+        if status in {"成功", "部分成功", "失败"}:
+            return status
+        if status:
             return "失败"
 
         value = self._row_measurement_mean(row)
diff --git a/alg/平整度、平面度、轮廓测量v1.5/test_pipeline_retention.py b/alg/平整度、平面度、轮廓测量v1.5/test_pipeline_retention.py
new file mode 100644
index 0000000..1a7e47a
--- /dev/null
+++ b/alg/平整度、平面度、轮廓测量v1.5/test_pipeline_retention.py
@@ -0,0 +1,62 @@
+from pathlib import Path
+from types import SimpleNamespace
+from unittest import TestCase
+from unittest.mock import Mock
+import tempfile
+
+import numpy as np
+
+from pipeline import MeasurementPipeline
+
+
+class CombinedContourRetentionTest(TestCase):
+    def make_pipeline(self, root: Path) -> MeasurementPipeline:
+        mesh = root / "mesh.ply"
+        pointcloud = root / "points.ply"
+        mesh.touch()
+        pointcloud.touch()
+
+        pipeline = MeasurementPipeline.__new__(MeasurementPipeline)
+        pipeline._resolve_contour_task_inputs = Mock(return_value=[{
+            "task_id": "1",
+            "sample_id": "sample",
+            "cad_path": None,
+            "mesh_path": mesh,
+            "pointcloud_path": pointcloud,
+        }])
+        pipeline._contour_task_save_path = Mock(return_value=root / "sample_contour.ply")
+        pipeline.run_contour = Mock(return_value={
+            "distance_stats": SimpleNamespace(max=0.5),
+            "distances": np.array([0.1, 0.2]),
+            "pointcloud": object(),
+        })
+        pipeline._contour_combined_save_path = Mock(return_value=root / "combined_contour.ply")
+        pipeline._save_contour_combined_pointcloud_stream = Mock()
+        pipeline._should_visualize = Mock(return_value=False)
+        return pipeline
+
+    def test_skips_combined_file_when_disabled(self) -> None:
+        with tempfile.TemporaryDirectory() as temp:
+            pipeline = self.make_pipeline(Path(temp))
+            result = pipeline.run_contour_batch(
+                all_tasks=True,
+                save_path=temp,
+                save_combined_visualization=False,
+                no_visualize=True,
+            )
+
+            self.assertIsNone(result["combined_visualization_output"])
+            pipeline._save_contour_combined_pointcloud_stream.assert_not_called()
+
+    def test_keeps_backward_compatible_combined_file_when_enabled(self) -> None:
+        with tempfile.TemporaryDirectory() as temp:
+            pipeline = self.make_pipeline(Path(temp))
+            result = pipeline.run_contour_batch(
+                all_tasks=True,
+                save_path=temp,
+                save_combined_visualization=True,
+                no_visualize=True,
+            )
+
+            self.assertEqual(result["combined_visualization_output"], Path(temp) / "combined_contour.ply")
+            pipeline._save_contour_combined_pointcloud_stream.assert_called_once()
diff --git a/scripts/backfill-flatness-suite-measurements.ts b/scripts/backfill-flatness-suite-measurements.ts
new file mode 100644
index 0000000..ca66348
--- /dev/null
+++ b/scripts/backfill-flatness-suite-measurements.ts
@@ -0,0 +1,275 @@
+#!/usr/bin/env node
+
+import { execFile as execFileCallback } from "node:child_process";
+import { promises as fs } from "node:fs";
+import path from "node:path";
+import { promisify } from "node:util";
+import { pathToFileURL } from "node:url";
+import mysql, { type Pool, type ResultSetHeader, type RowDataPacket } from "mysql2/promise";
+
+const execFile = promisify(execFileCallback);
+
+export type Measurement = Record<string, unknown>;
+
+export interface CandidateRow extends RowDataPacket {
+  id: number;
+  taskId: number;
+  measurements: unknown;
+  resultData: Record<string, unknown> | null;
+}
+
+export interface BackfillPlan {
+  row: CandidateRow;
+  resultJsonPath: string;
+  originalResultJson: Record<string, unknown>;
+  measurements: Measurement[];
+}
+
+export interface ExecutionSummary {
+  planned: number;
+  applied: number;
+  skipped: number;
+  failed: number;
+}
+
+export function mergeMeasurements<T extends Record<string, unknown>>(
+  value: T | null,
+  measurements: Measurement[],
+): T & { measurements: Measurement[] } {
+  return { ...(value ?? ({} as T)), measurements };
+}
+
+export async function executePlans(
+  plans: BackfillPlan[],
+  apply: boolean,
+  persist: (plan: BackfillPlan) => Promise<boolean>,
+): Promise<ExecutionSummary> {
+  const summary: ExecutionSummary = {
+    planned: plans.length,
+    applied: 0,
+    skipped: apply ? 0 : plans.length,
+    failed: 0,
+  };
+  if (!apply) return summary;
+
+  for (const plan of plans) {
+    try {
+      if (await persist(plan)) summary.applied += 1;
+      else summary.skipped += 1;
+    } catch (error) {
+      summary.failed += 1;
+      console.error(`[backfill] 结果 ${plan.row.id} 写入失败:`, error);
+    }
+  }
+  return summary;
+}
+
+function argumentValue(name: string): string | undefined {
+  const index = process.argv.indexOf(name);
+  return index >= 0 ? process.argv[index + 1] : undefined;
+}
+
+function timestamp(): string {
+  return new Date().toISOString().replace(/[:.]/g, "-");
+}
+
+async function parseMeasurements(
+  excelPath: string,
+  templatePath: string,
+  helperPath: string,
+): Promise<Measurement[]> {
+  const python = process.env.PYTHON3_PATH || "/usr/bin/python3";
+  const { stdout } = await execFile(
+    python,
+    [helperPath, "--excel", excelPath, "--template", templatePath],
+    { maxBuffer: 10 * 1024 * 1024 },
+  );
+  const parsed = JSON.parse(stdout) as unknown;
+  if (!Array.isArray(parsed) || parsed.length === 0) {
+    throw new Error("解析结果为空")
+  }
+  return parsed.filter((item): item is Measurement => !!item && typeof item === "object");
+}
+
+async function listCandidates(pool: Pool, resultId?: number): Promise<CandidateRow[]> {
+  const filters = [
+    "pluginId = 'flatness-suite'",
+    "status = 'completed'",
+    "(measurements IS NULL OR JSON_LENGTH(measurements) = 0)",
+  ];
+  const params: number[] = [];
+  if (resultId !== undefined) {
+    filters.push("id = ?");
+    params.push(resultId);
+  }
+  const [rows] = await pool.query<CandidateRow[]>(
+    `SELECT id, taskId, measurements, resultData
+       FROM task_analysis_results
+      WHERE ${filters.join(" AND ")}
+      ORDER BY id`,
+    params,
+  );
+  return rows;
+}
+
+async function buildPlans(
+  rows: CandidateRow[],
+  dataDir: string,
+  templatePath: string,
+  helperPath: string,
+): Promise<{ plans: BackfillPlan[]; skipped: number }> {
+  const plans: BackfillPlan[] = [];
+  let skipped = 0;
+  for (const row of rows) {
+    const resultDir = path.join(dataDir, "tasks", String(row.taskId), "analysis", String(row.id));
+    const resultJsonPath = path.join(resultDir, "result.json");
+    const excelPath = path.join(resultDir, "contour.xlsx");
+    try {
+      const [resultText] = await Promise.all([
+        fs.readFile(resultJsonPath, "utf8"),
+        fs.access(excelPath),
+      ]);
+      const resultJson = JSON.parse(resultText) as Record<string, unknown>;
+      const extra = resultJson.extra as Record<string, unknown> | undefined;
+      const resultDataExtra = row.resultData?.extra as Record<string, unknown> | undefined;
+      if (extra?.module !== "contour_batch" && resultDataExtra?.module !== "contour_batch") {
+        throw new Error("不是 contour_batch 批量结果")
+      }
+      const measurements = await parseMeasurements(excelPath, templatePath, helperPath);
+      plans.push({ row, resultJsonPath, originalResultJson: resultJson, measurements });
+      console.log(`[backfill] 结果 ${row.id}: 可回填 ${measurements.length} 条测量值`);
+    } catch (error) {
+      skipped += 1;
+      console.warn(`[backfill] 跳过结果 ${row.id}: ${error instanceof Error ? error.message : String(error)}`);
+    }
+  }
+  return { plans, skipped };
+}
+
+async function writeJsonAtomically(filePath: string, value: unknown): Promise<void> {
+  const temporaryPath = `${filePath}.backfill-${process.pid}.tmp`;
+  await fs.writeFile(temporaryPath, `${JSON.stringify(value, null, 2)}\n`, "utf8");
+  await fs.rename(temporaryPath, filePath);
+}
+
+async function writeBackup(plans: BackfillPlan[], dataDir: string): Promise<string> {
+  const backupDir = path.join(dataDir, "maintenance-backups");
+  await fs.mkdir(backupDir, { recursive: true });
+  const backupPath = path.join(backupDir, `flatness-suite-measurements-${timestamp()}.json`);
+  await writeJsonAtomically(backupPath, {
+    version: 1,
+    createdAt: new Date().toISOString(),
+    entries: plans.map((plan) => ({
+      id: plan.row.id,
+      taskId: plan.row.taskId,
+      resultJsonPath: plan.resultJsonPath,
+      measurements: plan.row.measurements,
+      resultData: plan.row.resultData,
+      resultJson: plan.originalResultJson,
+    })),
+  });
+  return backupPath;
+}
+
+async function persistPlan(pool: Pool, plan: BackfillPlan): Promise<boolean> {
+  const connection = await pool.getConnection();
+  const updatedResultJson = mergeMeasurements(plan.originalResultJson, plan.measurements);
+  const updatedResultData = mergeMeasurements(plan.row.resultData, plan.measurements);
+  try {
+    await connection.beginTransaction();
+    await writeJsonAtomically(plan.resultJsonPath, updatedResultJson);
+    const [result] = await connection.execute<ResultSetHeader>(
+      `UPDATE task_analysis_results
+          SET measurements = ?, resultData = ?
+        WHERE id = ?
+          AND pluginId = 'flatness-suite'
+          AND status = 'completed'
+          AND (measurements IS NULL OR JSON_LENGTH(measurements) = 0)`,
+      [JSON.stringify(plan.measurements), JSON.stringify(updatedResultData), plan.row.id],
+    );
+    if (result.affectedRows !== 1) {
+      await writeJsonAtomically(plan.resultJsonPath, plan.originalResultJson);
+      await connection.rollback();
+      return false;
+    }
+    await connection.commit();
+    return true;
+  } catch (error) {
+    await connection.rollback();
+    await writeJsonAtomically(plan.resultJsonPath, plan.originalResultJson).catch(() => undefined);
+    throw error;
+  } finally {
+    connection.release();
+  }
+}
+
+interface BackupManifest {
+  entries: Array<{
+    id: number;
+    resultJsonPath: string;
+    measurements: unknown;
+    resultData: unknown;
+    resultJson: unknown;
+  }>;
+}
+
+async function restoreBackup(pool: Pool, backupPath: string): Promise<void> {
+  const manifest = JSON.parse(await fs.readFile(backupPath, "utf8")) as BackupManifest;
+  if (!Array.isArray(manifest.entries)) throw new Error("备份清单格式无效")
+  for (const entry of manifest.entries) {
+    await writeJsonAtomically(entry.resultJsonPath, entry.resultJson);
+    await pool.execute(
+      "UPDATE task_analysis_results SET measurements = ?, resultData = ? WHERE id = ? AND pluginId = 'flatness-suite'",
+      [
+        entry.measurements == null ? null : JSON.stringify(entry.measurements),
+        entry.resultData == null ? null : JSON.stringify(entry.resultData),
+        entry.id,
+      ],
+    );
+    console.log(`[backfill] 已恢复结果 ${entry.id}`);
+  }
+}
+
+async function main(): Promise<void> {
+  const databaseUrl = process.env.DATABASE_URL;
+  if (!databaseUrl) throw new Error("缺少 DATABASE_URL")
+  const dataDir = process.env.DATA_DIR || "/data";
+  const algRoot = process.env.ALG_ROOT || "/alg";
+  const resultIdRaw = argumentValue("--result-id");
+  const resultId = resultIdRaw === undefined ? undefined : Number(resultIdRaw);
+  if (resultIdRaw !== undefined && !Number.isInteger(resultId)) throw new Error("--result-id 必须是整数")
+  const apply = process.argv.includes("--apply");
+  const restorePath = argumentValue("--restore");
+  const templatePath = path.join(algRoot, "平整度、平面度、轮廓测量v1.5", "excel", "测量结果.xlsx");
+  const helperPath = path.resolve("server/plugins/adapters/flatness_suite_measurements.py");
+  const pool = mysql.createPool(databaseUrl);
+
+  try {
+    if (restorePath) {
+      await restoreBackup(pool, restorePath);
+      return;
+    }
+    const rows = await listCandidates(pool, resultId);
+    const { plans, skipped } = await buildPlans(rows, dataDir, templatePath, helperPath);
+    let backupPath: string | undefined;
+    if (apply && plans.length > 0) {
+      backupPath = await writeBackup(plans, dataDir);
+      console.log(`[backfill] 写入前备份: ${backupPath}`);
+    }
+    const summary = await executePlans(plans, apply, (plan) => persistPlan(pool, plan));
+    summary.skipped += skipped;
+    console.log(`[backfill] ${apply ? "apply" : "dry-run"} 完成: ${JSON.stringify(summary)}`);
+    if (!apply && plans.length > 0) console.log("[backfill] 确认无误后追加 --apply 执行写入");
+    if (backupPath) console.log(`[backfill] 恢复命令: pnpm tsx scripts/backfill-flatness-suite-measurements.ts --restore ${backupPath}`);
+  } finally {
+    await pool.end();
+  }
+}
+
+const isMain = process.argv[1] && import.meta.url === pathToFileURL(path.resolve(process.argv[1])).href;
+if (isMain) {
+  main().catch((error) => {
+    console.error("[backfill] 失败:", error);
+    process.exitCode = 1;
+  });
+}
diff --git a/server/flatnessSuiteBackfill.test.ts b/server/flatnessSuiteBackfill.test.ts
new file mode 100644
index 0000000..c06f17d
--- /dev/null
+++ b/server/flatnessSuiteBackfill.test.ts
@@ -0,0 +1,62 @@
+import { describe, expect, it, vi } from "vitest";
+import {
+  executePlans,
+  mergeMeasurements,
+  type BackfillPlan,
+} from "../scripts/backfill-flatness-suite-measurements";
+
+function plan(id: number): BackfillPlan {
+  return {
+    row: { id, taskId: 24, measurements: [], resultData: { kind: "plugin", marker: "kept" } } as BackfillPlan["row"],
+    resultJsonPath: `/tmp/${id}/result.json`,
+    originalResultJson: { status: "completed", extra: { module: "contour_batch" } },
+    measurements: [{
+      name: "任务11 · 车钩面板的平整度",
+      value: null,
+      unit: "mm",
+      toleranceLabel: "0.5mm",
+      status: "fail",
+      passLabel: "成功",
+      warningLabel: "部分成功",
+      failLabel: "失败",
+    }],
+  };
+}
+
+describe("flatness-suite 历史回填", () => {
+  it("dry-run 不调用持久化", async () => {
+    const persist = vi.fn(async () => true);
+    const summary = await executePlans([plan(234)], false, persist);
+    expect(persist).not.toHaveBeenCalled();
+    expect(summary).toEqual({ planned: 1, applied: 0, skipped: 1, failed: 0 });
+  });
+
+  it("apply 由持久化层守卫实现幂等", async () => {
+    const applied = new Set<number>();
+    const persist = vi.fn(async (item: BackfillPlan) => {
+      if (applied.has(item.row.id)) return false;
+      applied.add(item.row.id);
+      return true;
+    });
+    expect((await executePlans([plan(234)], true, persist)).applied).toBe(1);
+    const second = await executePlans([plan(234)], true, persist);
+    expect(second.applied).toBe(0);
+    expect(second.skipped).toBe(1);
+  });
+
+  it("异常记录不阻断其他记录", async () => {
+    const persist = vi.fn(async (item: BackfillPlan) => {
+      if (item.row.id === 1) throw new Error("bad excel");
+      return true;
+    });
+    const summary = await executePlans([plan(1), plan(2)], true, persist);
+    expect(summary).toEqual({ planned: 2, applied: 1, skipped: 0, failed: 1 });
+  });
+
+  it("扩展字段完整进入 resultData 且保留已有字段", () => {
+    const item = plan(234);
+    const merged = mergeMeasurements(item.row.resultData, item.measurements);
+    expect(merged.marker).toBe("kept");
+    expect(merged.measurements).toEqual(item.measurements);
+  });
+});
diff --git a/server/plugins/adapters/flatness_suite_measurements.py b/server/plugins/adapters/flatness_suite_measurements.py
new file mode 100644
index 0000000..43daa6e
--- /dev/null
+++ b/server/plugins/adapters/flatness_suite_measurements.py
@@ -0,0 +1,246 @@
+"""flatness-suite 批量测量结果的模板关联与旧报表解析。"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import math
+import posixpath
+from pathlib import Path
+from typing import Any
+from zipfile import BadZipFile, ZipFile
+from xml.etree import ElementTree as ET
+
+
+_SHEET_NS = {"a": "http://schemas.openxmlformats.org/spreadsheetml/2006/main"}
+_REL_NS = "http://schemas.openxmlformats.org/officeDocument/2006/relationships"
+
+
+def _column_index(reference: str) -> int:
+    letters = "".join(character for character in reference if character.isalpha())
+    if not letters:
+        raise ValueError(f"无效的 Excel 单元格引用: {reference}")
+    result = 0
+    for character in letters.upper():
+        result = result * 26 + ord(character) - ord("A") + 1
+    return result
+
+
+def _cell_value(cell: ET.Element, shared_strings: list[str]) -> Any:
+    cell_type = cell.attrib.get("t")
+    if cell_type == "inlineStr":
+        return "".join(text.text or "" for text in cell.findall(".//a:t", _SHEET_NS))
+
+    value_node = cell.find("a:v", _SHEET_NS)
+    if value_node is None or value_node.text is None:
+        return ""
+    raw = value_node.text
+    if cell_type == "s":
+        try:
+            return shared_strings[int(raw)]
+        except (IndexError, ValueError) as exc:
+            raise ValueError("Excel 共享字符串索引无效") from exc
+    if cell_type in {"str", "e"}:
+        return raw
+    if cell_type == "b":
+        return raw == "1"
+    try:
+        number = float(raw)
+        return int(number) if number.is_integer() else number
+    except ValueError:
+        return raw
+
+
+def read_xlsx_rows(path: str | Path) -> tuple[str, list[list[Any]]]:
+    """只用标准库读取 xlsx 的第一个工作表。"""
+    workbook_path = Path(path)
+    if not workbook_path.is_file():
+        raise FileNotFoundError(f"Excel 文件不存在: {workbook_path}")
+
+    try:
+        with ZipFile(workbook_path) as xlsx:
+            names = set(xlsx.namelist())
+            shared_strings: list[str] = []
+            if "xl/sharedStrings.xml" in names:
+                shared_root = ET.fromstring(xlsx.read("xl/sharedStrings.xml"))
+                shared_strings = [
+                    "".join(text.text or "" for text in item.findall(".//a:t", _SHEET_NS))
+                    for item in shared_root.findall("a:si", _SHEET_NS)
+                ]
+
+            workbook_root = ET.fromstring(xlsx.read("xl/workbook.xml"))
+            first_sheet = workbook_root.find("a:sheets/a:sheet", _SHEET_NS)
+            if first_sheet is None:
+                raise ValueError(f"Excel 中没有工作表: {workbook_path}")
+            sheet_name = first_sheet.attrib.get("name", "Sheet1")
+            relation_id = first_sheet.attrib.get(f"{{{_REL_NS}}}id")
+            if not relation_id:
+                raise ValueError(f"Excel 工作表关系缺失: {workbook_path}")
+
+            rels_root = ET.fromstring(xlsx.read("xl/_rels/workbook.xml.rels"))
+            relation_targets = {
+                relation.attrib["Id"]: relation.attrib["Target"] for relation in rels_root
+            }
+            target = relation_targets.get(relation_id)
+            if not target:
+                raise ValueError(f"Excel 工作表文件关系缺失: {workbook_path}")
+            sheet_path = target.lstrip("/") if target.startswith("/") else posixpath.normpath(posixpath.join("xl", target))
+            if sheet_path not in names:
+                raise ValueError(f"Excel 工作表文件不存在: {sheet_path}")
+
+            sheet_root = ET.fromstring(xlsx.read(sheet_path))
+            sparse_rows: list[dict[int, Any]] = []
+            max_column = 0
+            for row in sheet_root.findall("a:sheetData/a:row", _SHEET_NS):
+                values: dict[int, Any] = {}
+                for cell in row.findall("a:c", _SHEET_NS):
+                    column = _column_index(cell.attrib.get("r", ""))
+                    values[column] = _cell_value(cell, shared_strings)
+                    max_column = max(max_column, column)
+                if values:
+                    sparse_rows.append(values)
+    except (BadZipFile, KeyError, ET.ParseError) as exc:
+        raise ValueError(f"无法解析 Excel 文件 {workbook_path}: {exc}") from exc
+
+    rows = [
+        [row.get(column, "") for column in range(1, max_column + 1)]
+        for row in sparse_rows
+    ]
+    if not rows:
+        raise ValueError(f"Excel 工作表为空: {workbook_path}")
+    return sheet_name, rows
+
+
+def _header_index(headers: list[Any], candidates: tuple[str, ...]) -> int:
+    normalized = [str(header).strip() for header in headers]
+    for candidate in candidates:
+        if candidate in normalized:
+            return normalized.index(candidate)
+    raise ValueError(f"Excel 缺少列: {'/'.join(candidates)}")
+
+
+def normalize_task_id(value: Any) -> str:
+    if isinstance(value, float) and value.is_integer():
+        return str(int(value))
+    text = str(value).strip()
+    return text[:-2] if text.endswith(".0") and text[:-2].isdigit() else text
+
+
+def _label(value: Any) -> str:
+    if value is None:
+        return ""
+    if isinstance(value, float) and value.is_integer():
+        return str(int(value))
+    return str(value).strip()
+
+
+def read_measurement_template(path: str | Path) -> dict[str, dict[str, str]]:
+    """读取任务号对应的项点名称与原始公差文本。"""
+    _, rows = read_xlsx_rows(path)
+    headers = rows[0]
+    task_column = _header_index(headers, ("任务号", "任务编号", "task_id"))
+    name_column = _header_index(headers, ("项点名称", "项目名称", "项目"))
+    tolerance_column = _header_index(headers, ("公差要求", "公差"))
+
+    result: dict[str, dict[str, str]] = {}
+    for row in rows[1:]:
+        task_id = normalize_task_id(row[task_column] if task_column < len(row) else "")
+        if not task_id:
+            continue
+        item_name = _label(row[name_column] if name_column < len(row) else "")
+        if not item_name:
+            raise ValueError(f"模板任务 {task_id} 缺少项点名称")
+        result[task_id] = {
+            "item_name": item_name,
+            "tolerance_label": _label(row[tolerance_column] if tolerance_column < len(row) else ""),
+        }
+    if not result:
+        raise ValueError(f"模板中没有测量任务: {path}")
+    return result
+
+
+def read_legacy_batch_rows(path: str | Path) -> list[dict[str, Any]]:
+    """读取旧版技术汇总表中的任务号、平均值和状态。"""
+    _, rows = read_xlsx_rows(path)
+    headers = rows[0]
+    task_column = _header_index(headers, ("任务号", "任务编号", "task_id"))
+    mean_column = _header_index(headers, ("平均值", "mean", "mean_dist", "average_distance"))
+    status_column = _header_index(headers, ("状态", "测量结论", "结论"))
+
+    result: list[dict[str, Any]] = []
+    for row in rows[1:]:
+        task_id = normalize_task_id(row[task_column] if task_column < len(row) else "")
+        if not task_id:
+            continue
+        result.append({
+            "task_id": task_id,
+            "mean": row[mean_column] if mean_column < len(row) else None,
+            "status": row[status_column] if status_column < len(row) else "",
+        })
+    if not result:
+        raise ValueError(f"旧版 Excel 中没有批量结果: {path}")
+    return result
+
+
+def _finite_float(value: Any) -> float | None:
+    try:
+        number = float(value)
+    except (TypeError, ValueError):
+        return None
+    return number if math.isfinite(number) else None
+
+
+def build_measurements(
+    rows: list[dict[str, Any]],
+    template_path: str | Path,
+) -> list[dict[str, Any]]:
+    template = read_measurement_template(template_path)
+    status_fields = {
+        "成功": ("pass", "成功"),
+        "部分成功": ("warning", "部分成功"),
+        "失败": ("fail", "失败"),
+    }
+    measurements: list[dict[str, Any]] = []
+    for row in rows:
+        task_id = normalize_task_id(row.get("task_id", ""))
+        if not task_id:
+            raise ValueError("批量结果包含空任务号")
+        template_row = template.get(task_id)
+        if template_row is None:
+            raise ValueError(f"模板中找不到任务 {task_id}")
+        raw_status = str(row.get("status", "")).strip()
+        if raw_status not in status_fields:
+            raise ValueError(f"任务 {task_id} 的状态无效: {raw_status or '<空>'}")
+        status, _ = status_fields[raw_status]
+        measurements.append({
+            "name": f"任务{task_id} · {template_row['item_name']}",
+            "value": _finite_float(row.get("mean")),
+            "unit": "mm",
+            "toleranceLabel": template_row["tolerance_label"],
+            "status": status,
+            "passLabel": "成功",
+            "warningLabel": "部分成功",
+            "failLabel": "失败",
+        })
+    return measurements
+
+
+def measurements_from_batch_result(result: dict[str, Any], template_path: str | Path) -> list[dict[str, Any]]:
+    rows = result.get("rows")
+    if not isinstance(rows, list):
+        raise ValueError("run_contour_batch 返回值缺少 rows")
+    return build_measurements(rows, template_path)
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(description="从旧版 flatness-suite Excel 构造通用 measurements")
+    parser.add_argument("--excel", required=True)
+    parser.add_argument("--template", required=True)
+    args = parser.parse_args()
+    measurements = build_measurements(read_legacy_batch_rows(args.excel), args.template)
+    print(json.dumps(measurements, ensure_ascii=False, allow_nan=False))
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())
diff --git a/server/plugins/adapters/flatness_suite_run.py b/server/plugins/adapters/flatness_suite_run.py
index 289b165..3df6e64 100644
--- a/server/plugins/adapters/flatness_suite_run.py
+++ b/server/plugins/adapters/flatness_suite_run.py
@@ -31,6 +31,7 @@ from _plugin_io import (
     write_error_result,
     write_result,
 )
+from flatness_suite_measurements import measurements_from_batch_result
 
 
 def _safe_float(v: object) -> float | None:
@@ -40,6 +41,25 @@ def _safe_float(v: object) -> float | None:
         return None
 
 
+def _config_bool(config: dict, key: str, default: bool = False) -> bool:
+    value = config.get(key, default)
+    if isinstance(value, bool):
+        return value
+    if isinstance(value, str):
+        return value.strip().lower() in {"1", "true", "yes", "on"}
+    return bool(value)
+
+
+def _cleanup_temporary_artifacts(out_dir: Path) -> None:
+    # 仅删除适配器自己创建、且不会写入 result.json 的固定临时文件。
+    # 不使用通配符，避免误删算法结果或用户要求保留的产物。
+    for name in ("_roi_cropped.ply",):
+        try:
+            (out_dir / name).unlink(missing_ok=True)
+        except OSError as exc:
+            print(f"[flatness-suite] 清理临时文件失败 {name}: {exc}", file=sys.stderr)
+
+
 def _mesh_cache_path(cad_path: str, mesh_deflection: float) -> Path:
     """根据 STEP 路径 + 文件指纹 + 三角化精度生成缓存路径。
 
@@ -191,12 +211,21 @@ def main() -> int:
         if use_batch_mode and module == "contour":
             emit_progress(0.20, "批量模式：从 datasets/Noise_PointCloud/ 读取所有面板点云")
             data_root = alg_dir / "datasets"
-            # 所选整车点云作为可视化背景；用户也可在 contextPointCloudPath 里手动指定
-            context_pc = (inputs.get("contextPointCloudPath")
-                          or pc_path_raw
-                          or None)
+            # combined_contour.ply 可能包含完整整车点云，单次可达十几 GB。
+            # 平台页面只使用各面板 *_contour.ply，因此默认从源头跳过组合文件；
+            # 仅在用户显式开启时才生成，并作为可下载产物登记。
+            keep_combined_contour = _config_bool(config, "keepCombinedContour", False)
+            context_pc = None
+            if keep_combined_contour:
+                context_pc = (config.get("contextPointCloudPath")
+                              or inputs.get("contextPointCloudPath")
+                              or pc_path_raw
+                              or None)
             save_dir = str(out_dir / "contour_batch")
             excel_path = str(out_dir / "contour.xlsx")
+            result_template_path = alg_dir / "excel" / "测量结果.xlsx"
+            if not result_template_path.is_file():
+                raise RuntimeError(f"内置测量结果模板不存在: {result_template_path}")
             res = pipeline.run_contour_batch(
                 all_tasks=True,
                 data_root=str(data_root),
@@ -205,20 +234,13 @@ def main() -> int:
                 triangle_chunk_size=512,
                 save_path=save_dir,
                 excel_output_path=excel_path,
+                result_template_path=str(result_template_path),
                 combined_context_pointcloud_path=context_pc,
+                save_combined_visualization=keep_combined_contour,
                 no_visualize=True,
             )
             emit_progress(0.95, "整理批量结果")
-            batch_stats = res.get("batch_stats") or {}
-            measurements: list[dict] = []
-            for task_id, task_res in (res.get("results") or {}).items():
-                s = task_res.get("distance_stats") or {}
-                if s:
-                    measurements.append({
-                        "name": f"任务{task_id} 轮廓偏差均值",
-                        "value": round(float(s.get("mean", 0)), 4),
-                        "unit": "mm", "status": "info",
-                    })
+            measurements = measurements_from_batch_result(res, result_template_path)
             # 各面板热力图 PLY 列表（不含背景点，全部是有意义的颜色）
             panel_plys = sorted([
                 f"contour_batch/{p.name}"
@@ -229,12 +251,19 @@ def main() -> int:
                 overlay3d = {"kind": "ply", "file": panel_plys[0], "files": panel_plys}
             else:
                 overlay3d = {"kind": "none"}
+            combined_rel = "contour_batch/combined_contour.ply"
+            result_files = [combined_rel] if keep_combined_contour and (out_dir / combined_rel).is_file() else []
             write_result(out_dir, {
                 "status": "completed",
-                "message": f"批量完成 {len(res.get('results', {}))} 个面板测量",
+                "message": f"批量完成 {len(res.get('rows') or [])} 个面板测量",
                 "measurements": measurements,
                 "overlay3d": overlay3d,
-                "extra": {"module": "contour_batch", "excelPath": "contour.xlsx"},
+                "files": result_files,
+                "extra": {
+                    "module": "contour_batch",
+                    "excelPath": "contour.xlsx",
+                    "combinedContourKept": bool(result_files),
+                },
             })
             return 0
 
@@ -430,6 +459,8 @@ def main() -> int:
             extra={"traceback": traceback.format_exc()},
         )
         return 1
+    finally:
+        _cleanup_temporary_artifacts(out_dir)
 
 
 if __name__ == "__main__":
diff --git a/server/plugins/adapters/test_flatness_suite_retention.py b/server/plugins/adapters/test_flatness_suite_retention.py
new file mode 100644
index 0000000..4c277a4
--- /dev/null
+++ b/server/plugins/adapters/test_flatness_suite_retention.py
@@ -0,0 +1,96 @@
+from pathlib import Path
+from tempfile import TemporaryDirectory
+from unittest import TestCase
+
+from flatness_suite_run import _cleanup_temporary_artifacts, _config_bool
+from flatness_suite_measurements import (
+    measurements_from_batch_result,
+    read_legacy_batch_rows,
+)
+
+
+TEMPLATE = (
+    Path(__file__).parents[3]
+    / "alg"
+    / "平整度、平面度、轮廓测量v1.5"
+    / "excel"
+    / "测量结果.xlsx"
+)
+
+
+class FlatnessSuiteRetentionConfigTest(TestCase):
+    def test_combined_contour_defaults_to_disabled(self) -> None:
+        self.assertFalse(_config_bool({}, "keepCombinedContour", False))
+
+    def test_explicit_boolean_or_string_can_enable_retention(self) -> None:
+        self.assertTrue(_config_bool({"keepCombinedContour": True}, "keepCombinedContour"))
+        self.assertTrue(_config_bool({"keepCombinedContour": "true"}, "keepCombinedContour"))
+        self.assertFalse(_config_bool({"keepCombinedContour": "false"}, "keepCombinedContour"))
+
+    def test_cleanup_removes_only_known_temporary_artifacts(self) -> None:
+        with TemporaryDirectory() as temp:
+            out_dir = Path(temp)
+            temporary = out_dir / "_roi_cropped.ply"
+            result = out_dir / "contour.ply"
+            temporary.touch()
+            result.touch()
+
+            _cleanup_temporary_artifacts(out_dir)
+
+            self.assertFalse(temporary.exists())
+            self.assertTrue(result.exists())
+
+
+class FlatnessSuiteBatchMeasurementTest(TestCase):
+    def test_uses_top_level_rows_instead_of_nested_sample_results(self) -> None:
+        result = {
+            "rows": [{"task_id": "11", "mean": 1.25, "status": "成功"}],
+            "results": {
+                "11": {
+                    "11": {"distance_stats": {"mean": 999}},
+                },
+            },
+        }
+
+        measurements = measurements_from_batch_result(result, TEMPLATE)
+
+        self.assertEqual(len(measurements), 1)
+        self.assertEqual(measurements[0]["name"], "任务11 · 车钩面板的平整度")
+        self.assertEqual(measurements[0]["value"], 1.25)
+        self.assertEqual(measurements[0]["toleranceLabel"], "0.5mm")
+
+    def test_maps_success_partial_success_failure_and_empty_value(self) -> None:
+        result = {
+            "rows": [
+                {"task_id": 11, "mean": 0.1, "status": "成功"},
+                {"task_id": 13, "mean": 0.2, "status": "部分成功"},
+                {"task_id": 17, "mean": None, "status": "失败"},
+            ],
+            "results": {},
+        }
+
+        measurements = measurements_from_batch_result(result, TEMPLATE)
+
+        self.assertEqual([item["status"] for item in measurements], ["pass", "warning", "fail"])
+        self.assertEqual(measurements[0]["passLabel"], "成功")
+        self.assertEqual(measurements[1]["warningLabel"], "部分成功")
+        self.assertEqual(measurements[2]["failLabel"], "失败")
+        self.assertIsNone(measurements[2]["value"])
+
+    def test_missing_template_has_explicit_error(self) -> None:
+        with self.assertRaisesRegex(FileNotFoundError, "Excel 文件不存在"):
+            measurements_from_batch_result(
+                {"rows": [{"task_id": 11, "mean": 0.1, "status": "成功"}]},
+                Path("missing-template.xlsx"),
+            )
+
+    def test_new_template_is_not_misread_as_legacy_summary(self) -> None:
+        with self.assertRaisesRegex(ValueError, "平均值"):
+            read_legacy_batch_rows(TEMPLATE)
+
+    def test_malformed_excel_has_explicit_error(self) -> None:
+        with TemporaryDirectory() as temp:
+            invalid = Path(temp) / "contour.xlsx"
+            invalid.write_text("not an xlsx", encoding="utf-8")
+            with self.assertRaisesRegex(ValueError, "无法解析 Excel"):
+                read_legacy_batch_rows(invalid)
diff --git a/server/plugins/manifests/flatness-suite.json b/server/plugins/manifests/flatness-suite.json
index 5b0d03d..32b1c72 100644
--- a/server/plugins/manifests/flatness-suite.json
+++ b/server/plugins/manifests/flatness-suite.json
@@ -3,7 +3,7 @@
   "name": "平整度 / 平面度 / 轮廓测量",
   "shortName": "flatness-suite",
   "description": "平整度 / 平面度 / 轮廓测量套件（v1.5）。两种模式：① 不选点云文件 → 批量模式，自动读取 alg/datasets/Noise_PointCloud/ 下全部面板点云（对应算法方标准测试流程）；② 选择点云文件 → 单点云模式，对所选点云做 contour/plane/curvature 测量。",
-  "version": "1.5.0",
+  "version": "1.5.1",
   "schemaVersion": 1,
   "category": "几何测量",
   "requiresPointCloud": false,
@@ -15,6 +15,13 @@
       "type": "boolean",
       "default": true
     },
+    {
+      "name": "keepCombinedContour",
+      "label": "保留组合轮廓点云（可能占用十几 GB）",
+      "type": "boolean",
+      "default": false,
+      "description": "默认关闭：页面使用各面板热力点云，不需要组合文件。仅在需要离线下载包含整车背景的 combined_contour.ply 时开启。"
+    },
     {
       "name": "module",
       "label": "测量模块",
@@ -47,10 +54,10 @@
     { "name": "roiZMax", "label": "ROI Z 最大值 (mm)", "type": "number", "default": null },
     {
       "name": "contextPointCloudPath",
-      "label": "整车背景点云路径（可视化合并用）",
+      "label": "整车背景点云路径（仅保留组合轮廓点云时使用）",
       "type": "string",
       "default": null,
-      "description": "可选。计算结果热力图将叠加到此点云上生成合并可视化文件。填整车配准后的完整点云路径（容器内路径），不填则只输出局部热力图。"
+      "description": "仅开启“保留组合轮廓点云”时生效。可填写容器内完整点云路径；留空时使用本次选择的点云作为背景。"
     }
   ],
   "runtime": {
diff --git a/server/plugins/types.ts b/server/plugins/types.ts
index d689132..aa98a62 100644
--- a/server/plugins/types.ts
+++ b/server/plugins/types.ts
@@ -106,11 +106,17 @@ export interface PluginProgressEvent {
 /** 测量项（GenericResultView 渲染） */
 export interface PluginMeasurement {
   name: string;
-  value: number;
+  /** 失败任务没有有效测量值时为 null，前端显示破折号。 */
+  value: number | null;
   unit?: string;
   tolerance?: number;
+  /** 无法安全数值化的原始公差表达式。 */
+  toleranceLabel?: string;
   /** pass / warning / fail / info */
   status?: 'pass' | 'warning' | 'fail' | 'info';
+  passLabel?: string;
+  warningLabel?: string;
+  failLabel?: string;
 }
 
 /** 误差统计卡片（与原 task_analysis_results.errorStats 兼容） */