diff --git a/docs/user/environment/1-atm-models/forecast.rst b/docs/user/environment/1-atm-models/forecast.rst
index 347fed281..17c8c3619 100644
--- a/docs/user/environment/1-atm-models/forecast.rst
+++ b/docs/user/environment/1-atm-models/forecast.rst
@@ -219,9 +219,17 @@ The following models are accepted:
 Let's see how to use Windy's data in RocketPy. First, we will set the location
 to EuRoC's launch area in Portugal.
 
+.. note::
 
-.. jupyter-execute::
+    These examples require access to Windy's forecast service. If the service is
+    unavailable, try another forecast source such as :ref:`open_meteo`.
+
+.. code-block:: python
+
+    from datetime import datetime, timedelta
+    from rocketpy import Environment
 
+    tomorrow = datetime.now() + timedelta(days=1)
     env_windy = Environment(
         date=tomorrow,
         latitude=39.3897,
@@ -238,7 +246,7 @@ you need reliable synoptic-scale forecasts several days ahead.
 
 We can use the ``ECMWF`` model from Windy.com. 
 
-.. jupyter-execute::
+.. code-block:: python
 
     env_windy.set_atmospheric_model(type="Windy", file="ECMWF")
     env_windy_ecmwf = env_windy
@@ -262,7 +270,7 @@ models when assessing forecast uncertainty.
 The ``GFS`` model is also available on Windy.com. This is the same model as
 described in the :ref:`global-forecast-system` section.
 
-.. jupyter-execute::
+.. code-block:: python
     
     env_windy.set_atmospheric_model(type="Windy", file="GFS")
     env_windy_gfs = env_windy
@@ -278,7 +286,7 @@ wind and temperature trends against GFS or ECMWF.
 
 The ICON model is a global weather forecasting model already available on Windy.com.
 
-.. jupyter-execute::
+.. code-block:: python
 
     env_windy.set_atmospheric_model(type="Windy", file="ICON")
     env_windy_icon = env_windy
diff --git a/docs/user/motors/genericmotor.rst b/docs/user/motors/genericmotor.rst
index 0706ffcba..30b404fa9 100644
--- a/docs/user/motors/genericmotor.rst
+++ b/docs/user/motors/genericmotor.rst
@@ -130,13 +130,14 @@ You can bypass the cache by setting ``no_cache=True``:
 .. note::
 
     This method performs network requests to the ThrustCurve API unless a cached
-    version exists. For automated testing or fully reproducible workflows, prefer
-    local `.eng` files or set ``no_cache=True`` explicitly.
+    version exists. If the service is unavailable or blocks the request, use
+    ``load_from_eng_file`` with a local ``.eng`` file instead. Local files are also
+    preferred for automated testing and fully reproducible workflows.
 
 Example
 -------
 
-.. jupyter-execute::
+.. code-block:: python
 
     from rocketpy.motors import GenericMotor
 
@@ -152,7 +153,7 @@ Using the no_cache option
 If you want to force RocketPy to ignore the cache and download a fresh copy
 every time, use:
 
-.. jupyter-execute::
+.. code-block:: python
 
     motor = GenericMotor.load_from_thrustcurve_api("M1670", no_cache=True)
     motor.info()