diff --git a/backend/app/api/indices.py b/backend/app/api/indices.py
index e78c2a2..febd6bb 100644
--- a/backend/app/api/indices.py
+++ b/backend/app/api/indices.py
@@ -113,7 +113,7 @@ def get_index_minute(
     repo = request.app.state.repo
     info = _index_info(repo, symbol)
     day = trade_date or date.today()
-    df = kline_sync.fetch_minute_single(symbol, day)
+    df = kline_sync.fetch_minute_single(symbol, day, asset_type="index")
     return {
         "symbol": symbol,
         "name": info.get("name"),
diff --git a/backend/app/api/kline.py b/backend/app/api/kline.py
index b4cfe1a..2018d6f 100644
--- a/backend/app/api/kline.py
+++ b/backend/app/api/kline.py
@@ -540,14 +540,38 @@ def get_minute_batch(request: Request, body: dict):
         start_time = datetime(trade_date.year, trade_date.month, trade_date.day, 9, 25, 0)
         end_time = datetime(trade_date.year, trade_date.month, trade_date.day, 15, 5, 0)
         lim = capset.limits(Cap.KLINE_MINUTE_BATCH)
-        live_df = kline_sync.sync_minute_batch(
-            incomplete,
-            start_time=start_time,
-            end_time=end_time,
-            batch_size=lim.batch if lim else None,
-            rpm=lim.rpm if lim else None,
-        )
-        if not live_df.is_empty():
+        # etf_set 已在上方获取, 直接复用 — 按 asset_type 拆分调用 sync_minute_batch
+        # (自定义源 / TickFlow 路由均依赖 asset_type 正确传递)
+        # 契约: 本端点只接受 stock/ETF (指数分钟K走 /api/index/minute 独立路径),
+        # 故两分支已覆盖全部 incomplete。若未来放开指数支持, 需额外加 index 分支
+        # 以避免被误路由为 stock。
+        stock_incomplete = [s for s in incomplete if s not in etf_set]
+        etf_incomplete = [s for s in incomplete if s in etf_set]
+        live_parts: list[pl.DataFrame] = []
+        if stock_incomplete:
+            df_s = kline_sync.sync_minute_batch(
+                stock_incomplete,
+                start_time=start_time,
+                end_time=end_time,
+                batch_size=lim.batch if lim else None,
+                rpm=lim.rpm if lim else None,
+                asset_type="stock",
+            )
+            if not df_s.is_empty():
+                live_parts.append(df_s)
+        if etf_incomplete:
+            df_e = kline_sync.sync_minute_batch(
+                etf_incomplete,
+                start_time=start_time,
+                end_time=end_time,
+                batch_size=lim.batch if lim else None,
+                rpm=lim.rpm if lim else None,
+                asset_type="etf",
+            )
+            if not df_e.is_empty():
+                live_parts.append(df_e)
+        if live_parts:
+            live_df = pl.concat(live_parts, how="diagonal_relaxed")
             for sym in incomplete:
                 sub = live_df.filter(pl.col("symbol") == sym).sort("datetime")
                 if not sub.is_empty():
@@ -594,7 +618,7 @@ def get_minute(
     if trade_date is None:
         # 本地无任何分钟K，尝试从 TickFlow 拉取当天
         trade_date = cn_today()
-        df = kline_sync.fetch_minute_single(symbol, trade_date)
+        df = kline_sync.fetch_minute_single(symbol, trade_date, asset_type=asset_type)
         price_limit = _get_price_limit_info(
             repo, symbol, trade_date, asset_type, stock_name,
         )
@@ -636,7 +660,7 @@ def get_minute(
         }
 
     # 本地不完整或无数据 → 从 TickFlow 实时拉取
-    live_df = kline_sync.fetch_minute_single(symbol, trade_date)
+    live_df = kline_sync.fetch_minute_single(symbol, trade_date, asset_type=asset_type)
     return {
         "symbol": symbol, "name": stock_name, "stock_info": stock_info,
         "date": str(trade_date), "rows": live_df.to_dicts(),
diff --git a/backend/app/data_providers/base.py b/backend/app/data_providers/base.py
index c17ca21..53b59f4 100644
--- a/backend/app/data_providers/base.py
+++ b/backend/app/data_providers/base.py
@@ -6,6 +6,7 @@ backtests stay data-source agnostic.
 """
 from __future__ import annotations
 
+from collections.abc import Callable
 from dataclasses import dataclass
 from datetime import datetime
 from typing import Literal, Protocol
@@ -57,8 +58,14 @@ class MarketDataProvider(Protocol):
         end_time: datetime | None,
         asset_type: AssetType,
         freq: str = "1m",
+        on_chunk_done: Callable[[int, int], None] | None = None,
     ) -> pl.DataFrame:
-        """Return normalized minute K rows. Implementations may return empty."""
+        """Return normalized minute K rows. Implementations may return empty.
+
+        on_chunk_done 契约: provider 实现内部以 2 参 (cur, total) 调用;
+        3 参 seg_label 适配由 kline_sync._try_custom_minute 包装层负责,
+        不应泄漏到 provider 契约层。
+        """
 
     def get_realtime(
         self,
diff --git a/backend/app/data_providers/custom/provider.py b/backend/app/data_providers/custom/provider.py
index 53b258e..1f80e91 100644
--- a/backend/app/data_providers/custom/provider.py
+++ b/backend/app/data_providers/custom/provider.py
@@ -3,6 +3,7 @@ from __future__ import annotations
 
 import logging
 import os
+from collections.abc import Callable
 from datetime import datetime
 from pathlib import Path
 from typing import Any
@@ -11,6 +12,7 @@ import httpx
 import polars as pl
 
 from app.config import settings
+from app.data_providers.base import AssetType
 from app.data_providers.custom.config import CustomSourceConfig, DatasetConfig
 from app.data_providers.custom.mapper import apply_transforms, datetime_payload, extract_rows, map_rows
 from app.data_providers.normalizer import normalize_adj_factors, normalize_daily
@@ -109,8 +111,9 @@ class GenericHTTPProvider:
         symbols: list[str],
         start_time: datetime | None,
         end_time: datetime | None,
-        asset_type: str = "stock",  # noqa: ARG002
-        on_chunk_done=None,
+        asset_type: AssetType = "stock",  # noqa: ARG002
+        freq: str = "1m",  # noqa: ARG002
+        on_chunk_done: Callable[[int, int], None] | None = None,
     ) -> pl.DataFrame:
         cfg = self._dataset("minute")
         frames: list[pl.DataFrame] = []
diff --git a/backend/app/data_providers/tickflow_provider.py b/backend/app/data_providers/tickflow_provider.py
index a086949..c69bd73 100644
--- a/backend/app/data_providers/tickflow_provider.py
+++ b/backend/app/data_providers/tickflow_provider.py
@@ -2,6 +2,7 @@
 from __future__ import annotations
 
 import logging
+from collections.abc import Callable
 from datetime import datetime
 
 import polars as pl
@@ -97,8 +98,9 @@ class TickFlowProvider:
         symbols: list[str],
         start_time: datetime | None,
         end_time: datetime | None,
-        asset_type: AssetType,  # noqa: ARG002
+        asset_type: AssetType = "stock",  # noqa: ARG002
         freq: str = "1m",  # noqa: ARG002
+        on_chunk_done: Callable[[int, int], None] | None = None,  # noqa: ARG002
     ) -> pl.DataFrame:
         # Existing minute sync remains in app.services.kline_sync for now.
         return pl.DataFrame()
diff --git a/backend/app/plugins/stocksdk/provider.py b/backend/app/plugins/stocksdk/provider.py
index 6922055..c3481a4 100644
--- a/backend/app/plugins/stocksdk/provider.py
+++ b/backend/app/plugins/stocksdk/provider.py
@@ -10,11 +10,13 @@ Original implementation by @forrany (PR #57), migrated to plugin architecture.
 from __future__ import annotations
 
 import logging
+from collections.abc import Callable
 from dataclasses import dataclass, field
 from datetime import datetime
 
 import polars as pl
 
+from app.data_providers.base import AssetType
 from app.data_providers.normalizer import normalize_adj_factors, normalize_daily
 from app.plugins.stocksdk import bridge
 from app.tickflow.rate_limits import chunked
@@ -135,13 +137,13 @@ class StockSDKProvider:
         symbols: list[str],
         start_time: datetime | None,
         end_time: datetime | None,
-        asset_type: str = "stock",  # noqa: ARG002
-        on_chunk_done=None,
-        freq: str = "5m",
+        asset_type: AssetType = "stock",  # noqa: ARG002
+        freq: str = "1m",
+        on_chunk_done: Callable[[int, int], None] | None = None,
     ) -> pl.DataFrame:
         if not symbols:
             return pl.DataFrame()
-        period = "".join(ch for ch in str(freq) if ch.isdigit()) or "5"
+        period = "".join(ch for ch in str(freq) if ch.isdigit()) or "1"
         logger.info("stock-sdk minute 拉取开始(%d symbols, period=%s)", len(symbols), period)
         frames: list[pl.DataFrame] = []
         chunks = chunked(symbols, _BATCH)
diff --git a/backend/app/services/kline_sync.py b/backend/app/services/kline_sync.py
index e2ba0db..2489718 100644
--- a/backend/app/services/kline_sync.py
+++ b/backend/app/services/kline_sync.py
@@ -9,10 +9,11 @@ from __future__ import annotations
 
 import logging
 from collections.abc import Callable
-from datetime import datetime, timedelta
+from datetime import date, datetime, timedelta
 
 import polars as pl
 
+from app.data_providers.base import AssetType
 from app.indicators.pipeline import filter_halt_days
 from app.market_time import cn_now
 from app.services import preferences
@@ -528,6 +529,56 @@ def _write_minute_partition(df: pl.DataFrame, minute_dir) -> int:
     return written
 
 
+def _try_custom_minute(
+    symbols: list[str],
+    start_time: datetime | None,
+    end_time: datetime | None,
+    asset_type: AssetType,
+    freq: str = "1m",
+    on_chunk_done: Callable[[int, int, str], None] | None = None,
+) -> tuple[pl.DataFrame | None, bool]:
+    """尝试从自定义分钟源拉取。返回 (df, should_fallback_to_tickflow)。
+
+    返回契约:
+      (None, True)   → 未配自定义源 / 未配 minute dataset / 自定义源异常 → 走 TickFlow
+      (df, False)    → 自定义源成功(含空 df) → 直接用, 不回退
+
+    降级策略 (C): 自定义源异常时无条件 fall through 到 TickFlow,
+    由 TickFlow 路径自身 try/except 兜底。Pro+ 用户 TickFlow 成功返回数据,
+    None 档用户 TickFlow 失败返回空。不显式判断 tier, 避免 #126 augmented
+    capability 逻辑干扰。
+
+    on_chunk_done 适配: 上层回调是 3 参 (cur, total, seg_label), provider
+    实现内部以 2 参 (cur, total) 调用。这里包装一层, provider 调 2 参时补
+    默认 seg_label="custom" 转发给上层, 保证进度展示不降级。
+    """
+    provider_name = preferences.get_minute_data_provider()
+    if provider_name == "tickflow":
+        return (None, True)
+    from app.data_providers import custom as custom_sources
+    if not custom_sources.provider_has_dataset(provider_name, "minute"):
+        return (None, True)
+    provider = custom_sources.get_provider(provider_name)
+
+    # 包装 on_chunk_done: provider 调 2 参 → 补 seg_label="custom" → 转发上层 3 参
+    wrapped_cb: Callable[[int, int], None] | None = None
+    if on_chunk_done is not None:
+        def _wrapped_cb(cur: int, total: int) -> None:
+            on_chunk_done(cur, total, "custom")
+        wrapped_cb = _wrapped_cb
+
+    try:
+        df = provider.get_minute(
+            symbols, start_time=start_time, end_time=end_time,
+            asset_type=asset_type, freq=freq, on_chunk_done=wrapped_cb,
+        )
+        return (df, False)
+    except Exception as e:  # noqa: BLE001
+        logger.warning("custom minute provider %s failed, falling back to TickFlow: %s",
+                       provider_name, e)
+        return (None, True)
+
+
 def sync_minute_batch(
     symbols: list[str],
     start_time: datetime | None = None,
@@ -538,6 +589,7 @@ def sync_minute_batch(
     on_chunk_done: Callable[[int, int, str], None] | None = None,
     segment_trading_days: int = 20,
     on_segment: Callable[[pl.DataFrame], None] | None = None,
+    asset_type: AssetType = "stock",
 ) -> pl.DataFrame:
     """批量拉取多股分钟 K。
 
@@ -555,16 +607,15 @@ def sync_minute_batch(
         不进入全局 out → 内存峰值从「全量」降到「单段」。适用于 sync_and_persist_minute。
         不传时 (如 get_minute_batch 的实时补拉) 保持原契约: 累积进 out 末尾一次性返回。
     """
-    # 自定义数据源分流: minute provider
-    provider_name = preferences.get_minute_data_provider()
-    if provider_name != "tickflow":
-        from app.data_providers import custom as custom_sources
-        if custom_sources.provider_has_dataset(provider_name, "minute"):
-            provider = custom_sources.get_provider(provider_name)
-            return provider.get_minute(
-                symbols, start_time=start_time, end_time=end_time, on_chunk_done=on_chunk_done,
-            )
-        # 未配置 minute → 回退 TickFlow
+    df, fallback = _try_custom_minute(
+        symbols, start_time=start_time, end_time=end_time,
+        asset_type=asset_type, freq="1m", on_chunk_done=on_chunk_done,
+    )
+    if not fallback:
+        # _try_custom_minute 成功时返回 (df, False), df 非 None;
+        # fallback=True 时才返回 (None, True)。故此处 df 必非 None。
+        # (旧版 `df or pl.DataFrame()` 触发 polars DataFrame __bool__ TypeError, 已修。)
+        return df if df is not None else pl.DataFrame()
 
     tf = get_client()
 
@@ -572,7 +623,7 @@ def sync_minute_batch(
     # 按 segment_trading_days 交易日分段 (交易日→自然日 ×7/5 换算, 含节假日余量)。
     seg_calendar_days = max(1, int(segment_trading_days * 7 / 5))
     SEG_CHUNK = timedelta(days=seg_calendar_days)
-    time_segments: list[tuple[datetime, datetime]] = []
+    time_segments: list[tuple[datetime | None, datetime | None]] = []
     if start_time and end_time:
         seg_start = start_time
         while seg_start < end_time:
@@ -589,10 +640,10 @@ def sync_minute_batch(
     # 段内累积: 每段拉完即 flush, 避免全量攒内存 (OOM 根因)
     seg_out: list[pl.DataFrame] = []
 
-    for seg_idx, (seg_start, seg_end) in enumerate(time_segments):
+    for seg_idx, (cur_start, cur_end) in enumerate(time_segments):
         # 当前的日期段描述 (供进度展示)
-        if seg_start and seg_end:
-            seg_label = f"{seg_start.strftime('%m-%d')}~{seg_end.strftime('%m-%d')}"
+        if cur_start and cur_end:
+            seg_label = f"{cur_start.strftime('%m-%d')}~{cur_end.strftime('%m-%d')}"
         else:
             seg_label = "最新"
         seg_total = len(time_segments)
@@ -601,11 +652,11 @@ def sync_minute_batch(
             sleep_between_batches(step, rpm)
             step += 1
             try:
-                if seg_start and seg_end:
+                if cur_start and cur_end:
                     raw = tf.klines.batch(
                         chunk, period="1m",
-                        start_time=_datetime_to_ms(seg_start),
-                        end_time=_datetime_to_ms(seg_end),
+                        start_time=_datetime_to_ms(cur_start),
+                        end_time=_datetime_to_ms(cur_end),
                         count=10000,
                         adjust="forward",
                         as_dataframe=True, show_progress=False,
@@ -737,7 +788,11 @@ def fetch_intraday_monitor_batch(
     return pl.concat(frames, how="diagonal_relaxed") if frames else pl.DataFrame()
 
 
-def fetch_minute_single(symbol: str, trade_date: date) -> pl.DataFrame:
+def fetch_minute_single(
+    symbol: str,
+    trade_date: date,
+    asset_type: AssetType = "stock",
+) -> pl.DataFrame:
     """实时拉取单股单日分钟 K(不写入本地)。优先自定义分钟源, 回退 TickFlow。"""
     from datetime import datetime
     start_time = datetime(trade_date.year, trade_date.month, trade_date.day, 9, 25, 0)
@@ -745,13 +800,13 @@ def fetch_minute_single(symbol: str, trade_date: date) -> pl.DataFrame:
 
     # 自定义数据源分流: 与 sync_minute_batch 一致, 配了自定义分钟源时走 custom provider,
     # 避免无 TickFlow Pro+ 权限的用户分时图首次打开(本地无数据)时补拉失败返回空。
-    provider_name = preferences.get_minute_data_provider()
-    if provider_name != "tickflow":
-        from app.data_providers import custom as custom_sources
-        if custom_sources.provider_has_dataset(provider_name, "minute"):
-            provider = custom_sources.get_provider(provider_name)
-            return provider.get_minute([symbol], start_time=start_time, end_time=end_time)
-        # 未配置 minute dataset → 回退 TickFlow
+    df, fallback = _try_custom_minute(
+        [symbol], start_time=start_time, end_time=end_time,
+        asset_type=asset_type, freq="1m",
+    )
+    if not fallback:
+        # 见 sync_minute_batch 同分支注释: df 在此必非 None。
+        return df if df is not None else pl.DataFrame()
 
     tf = get_client()
     try:
@@ -975,6 +1030,7 @@ def sync_and_persist_minute(
         on_chunk_done=on_chunk_done,
         segment_trading_days=segment_days,
         on_segment=_persist,
+        asset_type="stock",
     )
 
     if written_box[0] == 0:
diff --git a/backend/tests/test_minute_routing.py b/backend/tests/test_minute_routing.py
new file mode 100644
index 0000000..f4e6c10
--- /dev/null
+++ b/backend/tests/test_minute_routing.py
@@ -0,0 +1,312 @@
+"""自定义分钟数据源路由回归测试。
+
+对应设计文档 §4 测试矩阵 (docs/superpowers/specs/2026-07-18-minute-provider-unification-design.md)。
+
+覆盖三个阻断问题:
+1. stock-sdk 默认 freq 漂移 (5m → 1m)
+2. 自定义源异常直接 500 (无 try/except)
+3. 插件化路由重复 + asset_type 未透传
+
+mock 范式沿用 test_stocksdk_provider.py (monkeypatch 模块属性)。
+"""
+from __future__ import annotations
+
+from datetime import date, datetime
+from unittest.mock import MagicMock
+
+import httpx
+import polars as pl
+
+from app.plugins.stocksdk import provider as sp
+from app.plugins.stocksdk.provider import StockSDKProvider
+from app.services import kline_sync
+
+
+# ---------- 辅助 ----------
+
+def _mock_minute_df(symbol: str = "600519.SH") -> pl.DataFrame:
+    """构造非空分钟 K df, 用于 mock provider.get_minute 返回值。"""
+    return pl.DataFrame({
+        "symbol": [symbol],
+        "datetime": [datetime(2026, 1, 15, 9, 35, 0)],
+        "open": [100.0],
+        "high": [101.0],
+        "low": [99.5],
+        "close": [100.5],
+        "volume": [1000.0],
+        "amount": [100500.0],
+    })
+
+
+def _setup_custom_provider(monkeypatch, provider: object, has_dataset: bool = True) -> None:
+    """统一 mock 自定义分钟源路由前置: preferences + provider_has_dataset + get_provider。
+
+    - preferences.get_minute_data_provider → "mock_src"
+    - custom.provider_has_dataset → has_dataset
+    - custom.get_provider → provider
+    """
+    monkeypatch.setattr(
+        kline_sync.preferences,
+        "get_minute_data_provider",
+        lambda: "mock_src",
+    )
+    monkeypatch.setattr(
+        "app.data_providers.custom.provider_has_dataset",
+        lambda name, ds: has_dataset,
+    )
+    monkeypatch.setattr(
+        "app.data_providers.custom.get_provider",
+        lambda name: provider,
+    )
+
+
+# ---------- 测试 1: 自定义源成功返回 1 分钟 K ----------
+
+def test_custom_minute_provider_returns_1m_k(monkeypatch):
+    """§4 测试 1: 自定义源成功返回 1m K, 且 provider 收到 freq="1m"。"""
+    spy = MagicMock(return_value=_mock_minute_df())
+    mock_provider = MagicMock()
+    mock_provider.get_minute = spy
+    _setup_custom_provider(monkeypatch, mock_provider, has_dataset=True)
+
+    df, fallback = kline_sync._try_custom_minute(
+        ["600519.SH"],
+        datetime(2026, 1, 15, 9, 25, 0),
+        datetime(2026, 1, 15, 15, 5, 0),
+        asset_type="stock",
+    )
+
+    assert fallback is False
+    assert df is not None
+    assert not df.is_empty()
+    # spy 收到 freq="1m" 和 asset_type="stock"
+    spy.assert_called_once()
+    _, kwargs = spy.call_args
+    assert kwargs.get("freq") == "1m"
+    assert kwargs.get("asset_type") == "stock"
+
+
+# ---------- 测试 2: stock-sdk 收到 freq=1m → bridge job period="1" ----------
+
+def test_stocksdk_get_minute_receives_freq_1m(monkeypatch):
+    """§4 测试 2: StockSDKProvider.get_minute(freq="1m") → bridge job period == "1"。
+
+    bridge.mjs opMinute 用 String(period), 1m → "1"。
+    """
+    captured: dict = {}
+
+    def fake_run_job(job, timeout=None):
+        captured["job"] = job
+        # 返回空结果, 测试只验证 job.period
+        return {"ok": True, "op": job["op"], "rows": {}}
+
+    monkeypatch.setattr(sp.bridge, "run_job", fake_run_job)
+
+    StockSDKProvider().get_minute(
+        ["600519.SH"], None, None, freq="1m",
+    )
+
+    assert captured["job"]["op"] == "minute"
+    assert captured["job"]["period"] == "1"
+
+
+# ---------- 测试 3: 自定义源异常 + TickFlow 也失败 → 返回空 (非 500) ----------
+
+def test_custom_provider_exception_no_500(monkeypatch):
+    """§4 测试 3: 自定义源抛异常 + TickFlow 也失败,
+    fetch_minute_single / sync_minute_batch 返回空 df。
+    """
+    # 自定义源抛异常
+    mock_provider = MagicMock()
+    mock_provider.get_minute.side_effect = httpx.TimeoutException("timeout")
+    _setup_custom_provider(monkeypatch, mock_provider, has_dataset=True)
+
+    # mock get_client 返回 mock client, 其 klines.batch raise (TickFlow 也失败)
+    mock_tf = MagicMock()
+    mock_tf.klines.batch.side_effect = Exception("tickflow fail")
+    monkeypatch.setattr(kline_sync, "get_client", lambda: mock_tf)
+
+    # fetch_minute_single: 自定义源异常 → fall through → TickFlow 异常 → 返回空
+    df_single = kline_sync.fetch_minute_single(
+        "600519.SH", date(2026, 1, 15), asset_type="stock",
+    )
+    assert isinstance(df_single, pl.DataFrame)
+    assert df_single.is_empty()
+
+    # sync_minute_batch: 同一路径, 返回空
+    df_batch = kline_sync.sync_minute_batch(
+        ["600519.SH"],
+        start_time=datetime(2026, 1, 15, 9, 25, 0),
+        end_time=datetime(2026, 1, 15, 15, 5, 0),
+        asset_type="stock",
+    )
+    assert isinstance(df_batch, pl.DataFrame)
+    assert df_batch.is_empty()
+
+
+# ---------- 测试 4: 未配 minute dataset → 回退 TickFlow ----------
+
+def test_provider_without_minute_dataset_fallback(monkeypatch):
+    """§4 测试 4: provider_has_dataset 返回 False → (None, True) 回退 TickFlow。"""
+    mock_provider = MagicMock()
+    _setup_custom_provider(monkeypatch, mock_provider, has_dataset=False)
+
+    df, fallback = kline_sync._try_custom_minute(
+        ["600519.SH"], None, None, asset_type="stock",
+    )
+
+    assert fallback is True
+    assert df is None
+    # provider.get_minute 不应被调用 (回退决策在前)
+    mock_provider.get_minute.assert_not_called()
+
+
+# ---------- 测试 5: asset_type 透传到 provider ----------
+
+def test_asset_type_threaded_to_provider(monkeypatch):
+    """§4 测试 5: stock/etf/index asset_type 透传到 provider.get_minute。"""
+    spy = MagicMock(return_value=_mock_minute_df())
+    mock_provider = MagicMock()
+    mock_provider.get_minute = spy
+    _setup_custom_provider(monkeypatch, mock_provider, has_dataset=True)
+
+    # 三次调用不同 asset_type
+    kline_sync.fetch_minute_single("600519.SH", date(2026, 1, 15), asset_type="stock")
+    kline_sync.fetch_minute_single("510300.SH", date(2026, 1, 15), asset_type="etf")
+    kline_sync.fetch_minute_single("000001.SH", date(2026, 1, 15), asset_type="index")
+
+    # spy 被调 3 次, 每次收到对应 asset_type
+    assert spy.call_count == 3
+    received_assets = [call.kwargs.get("asset_type") for call in spy.call_args_list]
+    assert received_assets == ["stock", "etf", "index"]
+
+
+# ---------- 测试 6: 自定义源成功时不调 TickFlow ----------
+
+def test_custom_success_skips_tickflow(monkeypatch):
+    """§4 测试 6: fetch_minute_single 自定义源成功 → 不调 get_client。"""
+    expected_df = _mock_minute_df()
+    mock_provider = MagicMock()
+    mock_provider.get_minute.return_value = expected_df
+    _setup_custom_provider(monkeypatch, mock_provider, has_dataset=True)
+
+    # get_client 设为 spy, 若被调说明路由失败
+    get_client_spy = MagicMock(name="get_client_spy")
+    monkeypatch.setattr(kline_sync, "get_client", get_client_spy)
+
+    df = kline_sync.fetch_minute_single(
+        "600519.SH", date(2026, 1, 15), asset_type="stock",
+    )
+
+    # 返回的是 mock provider 的 df
+    assert df is expected_df
+    # TickFlow 路径未进入
+    get_client_spy.assert_not_called()
+
+
+# ---------- 测试 7: sync_minute_batch 自定义源成功直接返回 ----------
+
+def test_sync_minute_batch_custom_success_returns_directly(monkeypatch):
+    """§4 测试 7: sync_minute_batch 自定义源成功 → 直接 return, 不走 segment 逻辑。"""
+    expected_df = _mock_minute_df()
+    mock_provider = MagicMock()
+    mock_provider.get_minute.return_value = expected_df
+    _setup_custom_provider(monkeypatch, mock_provider, has_dataset=True)
+
+    get_client_spy = MagicMock(name="get_client_spy")
+    monkeypatch.setattr(kline_sync, "get_client", get_client_spy)
+
+    df = kline_sync.sync_minute_batch(
+        ["600519.SH"],
+        start_time=datetime(2026, 1, 15, 9, 25, 0),
+        end_time=datetime(2026, 1, 15, 15, 5, 0),
+        asset_type="stock",
+    )
+
+    # 返回 mock provider 的 df, 不走 segment
+    assert df is expected_df
+    get_client_spy.assert_not_called()
+
+
+# ---------- 测试 8: on_chunk_done 包装 (2参 → 3参补 seg_label='custom') ----------
+
+def test_on_chunk_done_wrapped_to_3_args(monkeypatch):
+    """on_chunk_done 包装: provider 内部以 2 参 (cur, total) 调用 →
+    上层 3 参 (cur, total, seg_label) spy 收到 seg_label='custom'。
+
+    设计文档 §2: 保证自定义源路径进度展示不降级 (与 TickFlow 路径 3 参回调对齐)。
+    """
+    upper_cb = MagicMock(name="upper_3arg_cb")
+
+    def provider_get_minute_side_effect(symbols, *, start_time, end_time,
+                                        asset_type, freq, on_chunk_done):
+        # 模拟 provider 实现内部以 2 参调用 on_chunk_done
+        # (如 GenericHTTPProvider/provider.py:127 / StockSDKProvider/provider.py:166)
+        if on_chunk_done is not None:
+            on_chunk_done(1, 3)
+        return _mock_minute_df()
+
+    mock_provider = MagicMock()
+    mock_provider.get_minute.side_effect = provider_get_minute_side_effect
+    _setup_custom_provider(monkeypatch, mock_provider, has_dataset=True)
+
+    df, fallback = kline_sync._try_custom_minute(
+        ["600519.SH"],
+        datetime(2026, 1, 15, 9, 25, 0),
+        datetime(2026, 1, 15, 15, 5, 0),
+        asset_type="stock",
+        on_chunk_done=upper_cb,
+    )
+
+    assert fallback is False
+    assert df is not None
+    # 上层 3 参 spy 被调用一次, 收到 (1, 3, "custom")
+    upper_cb.assert_called_once_with(1, 3, "custom")
+
+
+# ---------- 测试 9: get_minute_batch 按 asset_type 拆分调用 sync_minute_batch ----------
+
+def test_get_minute_batch_splits_stock_and_etf(monkeypatch):
+    """get_minute_batch 把 incomplete 拆成 stock/ETF 两组, 分别以
+    asset_type='stock'/'etf' 调用 sync_minute_batch, 结果 concat 返回。
+
+    覆盖 kline.py get_minute_batch 的双调用拼接逻辑 (本次提交改动量最大的部分)。
+    契约: 本端点只接受 stock/ETF (指数走 /api/index/minute), 故两分支覆盖全部 incomplete。
+    """
+    from app.api import kline as kline_api
+
+    # mock sync_minute_batch: stock 返回 df_s, etf 返回 df_e (不同 symbol 便于 concat 后 filter 验证)
+    def fake_sync(symbols, *, start_time, end_time, batch_size, rpm, asset_type):
+        if asset_type == "stock":
+            return _mock_minute_df(symbol="600519.SH")
+        if asset_type == "etf":
+            return _mock_minute_df(symbol="510300.SH")
+        return pl.DataFrame()
+    sync_spy = MagicMock(side_effect=fake_sync)
+    monkeypatch.setattr(kline_api.kline_sync, "sync_minute_batch", sync_spy)
+
+    # mock repo: ETF 集合含 510300.SH; 本地分钟K返回空 (强制走 incomplete 补拉)
+    mock_repo = MagicMock()
+    mock_repo.get_etf_symbol_set.return_value = {"510300.SH"}
+    mock_repo.get_minute_batch.return_value = pl.DataFrame()
+
+    # mock capset: 有权限, limits 返回 None (lim.batch 访问被 `if lim else` 守护)
+    mock_capset = MagicMock()
+    mock_capset.has.return_value = True
+    mock_capset.limits.return_value = None
+
+    mock_request = MagicMock()
+    mock_request.app.state.repo = mock_repo
+    mock_request.app.state.capabilities = mock_capset
+
+    body = {"symbols": ["600519.SH", "510300.SH"], "date": "2026-01-15"}
+    result = kline_api.get_minute_batch(mock_request, body)
+
+    # sync_minute_batch 被调 2 次, asset_type 分别为 stock 和 etf
+    assert sync_spy.call_count == 2
+    call_assets = sorted(call.kwargs.get("asset_type") for call in sync_spy.call_args_list)
+    assert call_assets == ["etf", "stock"]
+
+    # 两个 symbol 都在结果里 (concat 后按 symbol filter 命中)
+    assert "600519.SH" in result["data"]
+    assert "510300.SH" in result["data"]