diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 5b7ced194..cce0be404 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -38,6 +38,21 @@ jobs:
           csvstat < examples/dummy.csv
           in2csv -f csv < examples/dummy.csv
           sql2csv < examples/test.sql
+      - name: Read from pipe
+        run: |
+          printf 'a,b,c\n1,2,3' | csvclean
+          printf 'a,b,c\n1,2,3' | csvcut
+          printf 'a,b,c\n1,2,3' | csvformat
+          printf 'a,b,c\n1,2,3' | csvgrep -c1 -m1
+          printf 'a,b,c\n1,2,3' | csvjoin
+          printf 'a,b,c\n1,2,3' | csvjson
+          printf 'a,b,c\n1,2,3' | csvlook
+          printf 'a,b,c\n1,2,3' | csvsort
+          printf 'a,b,c\n1,2,3' | csvsql
+          printf 'a,b,c\n1,2,3' | csvstack
+          printf 'a,b,c\n1,2,3' | csvstat
+          printf 'a,b,c\n1,2,3' | in2csv -f csv
+          printf 'SELECT 1 AS a' | sql2csv
       - env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         run: coveralls --service=github
diff --git a/csvkit/utilities/in2csv.py b/csvkit/utilities/in2csv.py
index 57bc75e68..9248a51f5 100644
--- a/csvkit/utilities/in2csv.py
+++ b/csvkit/utilities/in2csv.py
@@ -64,6 +64,7 @@ def add_arguments(self):
             '-I', '--no-inference', dest='no_inference', action='store_true',
             help='Disable type inference (and --locale, --date-format, --datetime-format) when parsing CSV input.')
 
+    # This is called only from open_excel_input_file(), but is a separate method to use caching.
     @functools.lru_cache
     def stdin(self):
         return sys.stdin.buffer.read()