> ## Documentation Index
> Fetch the complete documentation index at: https://developer.onecodex.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Definining parameters (arguments)

## For Shell script/Docker workflows

When creating your workflow, you will see a section for Arguments to the right of your script. While in a draft state, you can add/remove arguments from this section. For each argument, we will provide fields for the following details:

* **Name:** Provide a name for your argument. When launching the analysis for a sample via the API, you can pass this name and its value. To access the argument in your script, we will prepend `ARGS_` to the name, and we will capitalize the name. For instance, if you named your argument `your_argument_1`, you can access this in the script as `$ARGS_YOUR_ARGUMENT_1`.
* **Type:** Choose from a dropdown menu for the value type, such as `boolean`, `number`, `integer`, `regex`, `string`.
* **Required:** Check the box to determine if this argument is a required argument. If you check the box, you must provide a value at the time of launch.
* **Description:** An optional field to provide details on what this argument is for.

## For Nextflow workflows

Nextflow workflows use the `nextflow_schema.json` file, which must be stored in the root directory of your GitHub repository, to define the parameters that will be used by the workflow. These arguments will be visible on the workflow edit/details page, where you can additionally set whether the argument is required or not.

To access the argument in your script, we will prepend `ARGS_` to the name, and we will capitalize the name. For instance, if you named your argument `your_argument_1`, you can access this in the script as `$ARGS_YOUR_ARGUMENT_1`.

For Nextflow workflows, your argument values will be written to a file named `input_params.json`, which you will pass to your `nextflow run` command, such as:

```
nextflow run "${REPOSITORY_DIR}/main.nf" \
  -params-file input_params.json \
  --outdir output-$OCX_ANALYSIS_UUID
```

## Setting Argument Values at Launch

<Info>
  Any parameters that are marked as required must be passed when launching the workflow. If required arguments have default values, these default values will automatically be populated in the web app, but they are not assumed when launching via the API, and must be provided, even if default values are being used.
</Info>

To override the default parameters for a workflow run, or to set values for parameters that do not have defaults, you can:

1. Pass the values in the confirmation screen when launching via the web app
2. If launching the workflow via API, you can provide the arguments directly in the command, or store them in a .json file. For instance, if you saved the below as `my_workflow_arguments.json`:

<CodeGroup>
  ```json json theme={null}
  {
      "sample":"the UUID of the sample you want to run the analysis on",
      "job_args": {
        	"your_argument_1": "arg_1_value",
          "your_argument_2": "arg_2_value",
      }
  }
  ```
</CodeGroup>

* then you can run the workflow, with the above arguments on the sample provided in the JSON, with the below command

<CodeGroup>
  ```sh curl theme={null}
  curl -u $ONE_CODEX_API_KEY: -H "Content-Type: application/json" -d @my_workflow_arguments.json -X POST https://app.onecodex.com/api/v1/jobs/[workflow_UUID]/run
  ```
</CodeGroup>

* where `$ONE_CODEX_API_KEY` is your API key,
* `workflow_UUID` is the unique identifier for the workflow that you want to run on the sample

3. Alternatively, you can pass the arguments directly to the cURL command, such as:

```
curl -u $ONE_CODEX_API_KEY: -H "Content-Type: application/json" -d {"sample":"[sample_UUID]", "job_args":{"your_argument_1": "arg_1_value","your_argument_2": "arg_2_value",}} -X POST https://app.onecodex.com/api/v1/jobs/[workflow_UUID]/run
```

***

<Icon icon="forward" iconType="duotone" size={18} /> **What’s next?** Now that you've defined your parameters, learn how to set them.