diff --git a/src/loaders/curl/README.md b/src/loaders/curl/README.md
new file mode 100644
index 0000000..e3feb6c
--- /dev/null
+++ b/src/loaders/curl/README.md
@@ -0,0 +1,105 @@
+# Curl loader
+
+## Configuration
+
+To define load behavior, create a config.json file.
+
+For example:
+
+If you want to test how your API endpoint on the frontend service handles concurrent traffic, you can list it multiple times in the `urls` array and adjust the `sleep` and `wait` values accordingly to simulate the desired load pattern.
+
+```json
+{
+  "sleep": 2,
+  "wait": 5,
+  "urls": [
+    "http://frontend/upload",
+    "http://frontend/upload",
+    "http://frontend/upload"
+  ]
+}
+```
+
+Each of the following parameters must be specified to control how the load is executed:
+
+| field          | type           | description                                |
+| -------------- | -------------- | ------------------------------------------ |
+| Sleep          | number         | seconds to sleep between each request loop |
+| Wait           | number         | initial wait time before the first request |
+| URLs           | array          | list of service URLs to target             |
+
+## Behavior Summary
+
+- **Initial Delay**: Waits for the duration specified in wait (in seconds) before initiating any requests.
+- **Request Execution**: Sends repeated HTTP GET requests to each URL in the list, automatically appending a `?unique_session_id=<uuid>` parameter to ensure each request is distinct.
+- **Loop Interval**: Pauses for the duration specified in sleep (in seconds) between each full cycle of requests to all URLs.
+
+## Dependencies
+
+The script requires the following in its runtime environment:
+
+- [`curl`](https://curl.se/)
+- [`jq`](https://jqlang.org/)
+- `uuidgen` (part of uuid-runtime or util-linux)
+
+## Getting Started with Docker
+
+### Docker
+
+Build the Docker image:
+
+```bash
+   docker build -t curl-loader
+```
+
+Run the container:
+
+```bash
+   docker run --rm -v $(pwd)/config.json:/config.json curl-loader
+```
+
+### 2. Docker Compose Integration
+
+Use this block in your config.yaml:
+
+```yaml
+loaders: 
+  user-1:   
+   type: curl   
+   wait: 0   
+   sleep: 2   
+   urls:     
+     - http://frontend/upload     
+     - http://frontend/upload     
+     - http://frontend/upload
+```
+
+Generate the Docker Compose config:
+
+```bash
+  docker run --rm -v ${PWD}:/mnt \ 
+  ghcr.io/cisco-open/app-simulator-generators-docker-compose  
+  --config /mnt/config.yaml \ 
+  --output /mnt/docker-compose.yaml
+```
+
+### 3. Example Scenario
+
+Here is an example of `docker-compose.yaml` setup:
+
+```yaml
+services: 
+  frontend:   
+    image: ghcr.io/cisco-open/app-simulator-services-java:edge   
+    ports:     
+      - "3000:80" 
+  user-1:   
+    image: ghcr.io/cisco-open/app-simulator-loaders-curl:edge
+```
+
+After running `docker compose up`, the loader will continuously simulate user-traffic to frontend.
+
+## Tracing with Jaeger
+
+Combine this loader with OpenTelemetry instrumentation in your services and Jaeger to visualize traces flowing through your architecture.
+See the [Observability with OpenTelemetry](../../../docs/tutorial/5-observability-with-opentelemetry.md) for details.