> For the complete documentation index, see [llms.txt](https://seanime.gitbook.io/seanime-extensions/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://seanime.gitbook.io/seanime-extensions/plugins/apis/system/commands.md).

# Commands

{% hint style="warning" %}
As of Seanime `v3.8.0`, if the user enables Extension Secure Mode, command execution can prompt for confirmation.

If the user rejects the prompt, the command call throws. If prompts cannot be shown, such as during app startup before a UI client is connected, the command call fails immediately.
{% endhint %}

## Permissions

By default, Seanime disallows running commands, you must manually defines the command and arguments your plugin will want to run using `commandScopes` .

### Example

<pre class="language-json"><code class="lang-json">{
  // ...
  "plugin": {
    "permissions": {
      "allow": {
        "readPaths": ["$DOWNLOAD"],
        "writePaths": ["$DOWNLOAD"],
        "commandScopes": [
          {
            "command": "ls",
<strong>            "args": [{ "value": "-la" }, { "validator": "$PATH" }]
</strong>          },
          {
            "command": "grep",
<strong>            "args": [{ "value": "Hello" }, { "validator": "$PATH" }]
</strong>          },
          {
            "command": "sort",
            "args": []
          },
          {
            "command": "echo",
<strong>            "args": [{ "validator": "$ARGS" }]
</strong>          },
          {
            "command": "open",
<strong>            "args": [{ "validator": "^https?://.*$" }]
</strong>          }
        ]
      }
    }
  }
}
</code></pre>

This example shows:

* The `ls` command can be executed with the `-la` argument followed by a valid file/directory path `$PATH`. This path must be in the allow list for `write` . `$PATH` is an alternative to writing the regex.
* The `grep` command is allowed with the "Hello" argument and a similar path validation.
* The `sort` command is permitted without any additional arguments.
* The `echo` command is allowed with any argument or list of arguments.
* The `open` command is allowed with any valid URLs

## Command (sync)

The code below shows how to run a command with the caveat that this approach will block the plugin's UI context thread until the command finishes running.

```typescript
const tempDir = $os.tempDir();

try {
	// Create a command to list files
	const cmd = $os.cmd("ls", "-la", tempDir);
	
	// Set up stdout capture
	const stdoutPipe = cmd.stdoutPipe();
	
	// Start the command
	cmd.start();
	
	// Read the output
	const output = $io.readAll(stdoutPipe);
	console.log($toString(output));
	
	// Wait for the command to complete
	cmd.wait();
	
	// Check exit code
	const exitCode = cmd.processState.exitCode();
	console.log("Command exit code:", exitCode); // Command exit code: 0
} catch (e) {
	console.log("Command execution error:", e.message);
}
```

## Command (async)

If you need to run a command without blocking the plugin's UI context thread, you should use `$osExtra.asyncCmd` .

{% hint style="info" %}
Do not use both sync and async commands in the same plugin as this can cause some data race issues.
{% endhint %}

```typescript
const tempDir = $os.tempDir();

try {
	// Create a command to list files
	const cmd = $osExtra.asyncCmd("ls", "-la", tempDir);
	
	// The callback function will fire for each new line of the stdout, stderr
	// and when the command finishes executing.
	cmd.run((data, err, exitCode, signal) => {
		// Stdout
		if (data) {
			console.log("Data:", $toString(data));
		}
		// Stderr
		if (err) {
			console.log("Error:", $toString(err));
		}
		// Command exited
		if (exitCode !== undefined) {
			console.log("Exited:", exitCode, signal);
		}
	});
	
	console.log("Doesn't wait for the command to finish!")
	
	// You still have access to the underlying command
	const _cmd = cmd.getCommand()
	
	
} catch (e) {
	console.log("Command execution error:", e.message);
}
```