diff --git a/tests/test_alpha_vantage_hardening.py b/tests/test_alpha_vantage_hardening.py
index d82ea8dc3..494f2bac6 100644
--- a/tests/test_alpha_vantage_hardening.py
+++ b/tests/test_alpha_vantage_hardening.py
@@ -3,7 +3,8 @@
 Regressions for #990 (no request timeout -> can hang), #991 (invalid-key
 responses mislabeled as rate limits and silently treated as transient), and
 #1115 (fundamentals look-ahead filter never ran because the payload is a JSON
-string, not a dict).
+string, not a dict), and the date trim that keeps post-end_date bars out of a
+historical run.
 """
 import json
 
@@ -11,6 +12,7 @@ import pytest
 
 import tradingagents.dataflows.alpha_vantage_common as av
 import tradingagents.dataflows.alpha_vantage_fundamentals as avf
+import tradingagents.dataflows.alpha_vantage_stock as avs
 
 
 class _FakeResponse:
@@ -94,3 +96,40 @@ def test_fundamentals_no_curr_date_passes_through(monkeypatch):
 def test_fundamentals_non_json_body_unchanged(monkeypatch):
     monkeypatch.setattr(avf, "_make_api_request", lambda fn, params: "not-json")
     assert avf.get_cashflow("AAPL", curr_date="2024-01-01") == "not-json"
+
+
+# ---------------------------------------------------------------------------
+# Date trim (see the rationale on the unguarded trim in alpha_vantage_common)
+# ---------------------------------------------------------------------------
+
+_DAILY_CSV = (
+    "timestamp,open,high,low,close,volume\n"
+    "2024-05-13,1,1,1,1,10\n"   # after end_date -> must never be served
+    "2024-05-10,1,1,1,1,10\n"
+    "2024-05-09,1,1,1,1,10\n"
+)
+
+
+@pytest.mark.unit
+def test_stock_data_is_trimmed_to_the_requested_window(monkeypatch):
+    monkeypatch.setattr(avs, "_make_api_request", lambda *a, **k: _DAILY_CSV)
+    out = avs.get_stock("IBM", "2024-05-09", "2024-05-10")
+    assert "2024-05-10" in out and "2024-05-09" in out
+    assert "2024-05-13" not in out, "bar after end_date leaked into the window"
+
+
+@pytest.mark.unit
+def test_unparseable_body_is_never_served_untrimmed(monkeypatch):
+    """The trim used to swallow the failure and return the whole body, putting
+    bars after end_date into a backtest. It must raise instead."""
+    monkeypatch.setattr(avs, "_make_api_request",
+                        lambda *a, **k: "timestamp,close\nnot-a-date,1\n")
+
+    with pytest.raises(ValueError):
+        avs.get_stock("IBM", "2024-05-09", "2024-05-10")
+
+
+@pytest.mark.unit
+def test_empty_body_still_passes_through(monkeypatch):
+    monkeypatch.setattr(avs, "_make_api_request", lambda *a, **k: "")
+    assert avs.get_stock("IBM", "2024-05-09", "2024-05-10") == ""
diff --git a/tradingagents/dataflows/alpha_vantage_common.py b/tradingagents/dataflows/alpha_vantage_common.py
index 237ff54ce..b35feccd9 100644
--- a/tradingagents/dataflows/alpha_vantage_common.py
+++ b/tradingagents/dataflows/alpha_vantage_common.py
@@ -128,24 +128,18 @@ def _filter_csv_by_date_range(csv_data: str, start_date: str, end_date: str) ->
     if not csv_data or csv_data.strip() == "":
         return csv_data
 
-    try:
-        # Parse CSV data
-        df = pd.read_csv(StringIO(csv_data))
+    # Deliberately unguarded: TIME_SERIES_DAILY_ADJUSTED returns the full series
+    # up to today, so this trim is the only thing keeping bars after end_date out
+    # of a historical run. Returning the untrimmed body on failure would leak
+    # future prices, so a parse failure propagates and the caller fails closed.
+    df = pd.read_csv(StringIO(csv_data))
 
-        # Assume the first column is the date column (timestamp)
-        date_col = df.columns[0]
-        df[date_col] = pd.to_datetime(df[date_col])
+    # Assume the first column is the date column (timestamp)
+    date_col = df.columns[0]
+    df[date_col] = pd.to_datetime(df[date_col])
 
-        # Filter by date range
-        start_dt = pd.to_datetime(start_date)
-        end_dt = pd.to_datetime(end_date)
+    start_dt = pd.to_datetime(start_date)
+    end_dt = pd.to_datetime(end_date)
+    filtered_df = df[(df[date_col] >= start_dt) & (df[date_col] <= end_dt)]
 
-        filtered_df = df[(df[date_col] >= start_dt) & (df[date_col] <= end_dt)]
-
-        # Convert back to CSV string
-        return filtered_df.to_csv(index=False)
-
-    except Exception as e:
-        # If filtering fails, return original data with a warning
-        print(f"Warning: Failed to filter CSV data by date range: {e}")
-        return csv_data
+    return filtered_df.to_csv(index=False)