diff --git a/docs/developer-guide/test-and-debug/ksqldb-testing-tool.md b/docs/developer-guide/test-and-debug/ksqldb-testing-tool.md
index 8fdb0c437b5c..f594d9e0e236 100644
--- a/docs/developer-guide/test-and-debug/ksqldb-testing-tool.md
+++ b/docs/developer-guide/test-and-debug/ksqldb-testing-tool.md
@@ -86,6 +86,23 @@ the previous test:
 }
 ```
 
+#### Generating input files automagically from existing topic data
+
+You can use [`kafkacat`](https://github.com/edenhill/kafkacat) and 
+[`jq`](https://stedolan.github.io/jq/) in combination to create an input file 
+based on data already in a Kafka topic: 
+
+```bash
+kafkacat -b broker:29092 -t my_topic -C -e -J | \
+  jq --slurp '{inputs:[.[]|{topic:.topic,timestamp: .ts, key: .key, value: .payload|fromjson}]}' \
+  > input.json
+```
+
+Replace `broker:29092` with your broker host and port, and `my_topic` with the 
+name of your topic. You can limit how many messages are written to the file by 
+adding a `-c` flag to the `kafkacat` statement—for example, `-c42` would write 
+the first 42 messages from the topic. 
+
 ### Output File
 
 The output file is a JSON file with an array field named `outputs`.
@@ -135,6 +152,17 @@ a window field:
 Currently, in the input files you can have only records with session
 window types.
 
+#### Generating output files automagically from existing topic data
+
+You can use a similar approach to that shown above for generating input files, 
+but be aware that it won't handle time windows in cases where these are 
+required. In the case of aggregates it will often only have the final value, 
+and not every intermediate value that is generated by the testing tool due to 
+buffering (see discussion in "Input Consumption" below). 
+
+Running the testing tool
+------------------------
+
 The testing tool indicates the success or failure of a test by
 printing the corresponding message. The following is the result of a
 successful test:
@@ -146,9 +174,12 @@ ksql-test-runner -s statements.sql -i input.json -o output.json
 Your output should resemble:
 
 ```
-Test passed!
+>>> Test passed!
 ```
 
+Note that the tool may also write verbose log output to the terminal too, in 
+which case you may need to page through it to locate the test status message.
+
 If a test fails, the testing tool will indicate the failure along with
 the cause. Here is an example of the output for a failing test: