use auto numerated list

docs/source/noaa_csp.rst

@@ -327,7 +327,7 @@ Running the Global Workflow
 Assume you have a AWS cluster running, after login to the cluster through `ssh` from your laptop terminal,
 or access the cluster from your web terminal, one can start clone, compile, and run global-workflow.
 
-1. clone global-workflow(assume you have setup access to githup)::
+#. clone global-workflow(assume you have setup access to githup)::
 
 .. code-block:: console
 
@@ -336,7 +336,7 @@ or access the cluster from your web terminal, one can start clone, compile, and
      #or the develop form at EPIC:
      git clone --recursive git@github.com:NOAA-EPIC/global-workflow-cloud.git global-workflow-cloud
 
-2. compile global-workflow::
+#. compile global-workflow::
 
 .. code-block:: console
 
@@ -345,7 +345,8 @@ or access the cluster from your web terminal, one can start clone, compile, and
      build_all.sh   # or similar command to compile for gefs, or others.
      link_workflow.sh  # after build_all.sh finished successfully
 
-3. As users may define a very small cluster as controller, one may use a script similar to this to compile in compute node::
+#. As users may define a very small cluster as controller, one may use the script below to compile in compute node.
+   Save the this script in a file, say, com.slurm, and submit this job with command "sbatch com.slurm"::
 
 .. code-block:: console
 
@@ -371,9 +372,7 @@ or access the cluster from your web terminal, one can start clone, compile, and
 
      link_workflow.sh
 
-Save the above lines in a file, say, com.slurm, and submit this job with command "sbatch com.slurm"
-
-4. run global-workflow C48 ATM test case (assume user has /lustre filesystem attached)::
+#. run global-workflow C48 ATM test case (assume user has /lustre filesystem attached)::
 
 .. code-block:: console