diff --git a/README.md b/README.md
new file mode 100644
index 0000000..06a5132
--- /dev/null
+++ b/README.md
@@ -0,0 +1,15 @@
+# ismdata
+
+Reduce [Umadump](https://github.com/Werseter/umadump/releases)'s idle single mode (Independent Training) outputs to CSV for analysis.
+
+Open Umadump's idle_single_mode folder, select all files with Ctrl+A, and drag and drop the files onto ismdata.exe to create idle_single_mode.csv.
+
+Alternatively, run ismdata in the command line with the filenames to process as arguments.
+When using it this way, `-o <output.csv>` can be provided to change the output file.
+Stdout is not supported.
+
+## CSV Contents
+
+The output CSV has a header row followed by eight rows per input career dump.
+Each row gives the stat gains from one respective source, along with distinctive information per career: scenario, full deck, race count and wins, and negative and positive conditions.
+The `GainSource` column is `Events`, `Inspiration`, or a support card ID.