ASF 支援一些能夠影響程式運行時的命令列參數。 高級用戶可使用這些參數以定義程式運行方式。 與ASF.json 配置文件的預設方式相比，命令列參數可用於核心初始化（例如--path）、平台特定設置（例如--system-required）或敏感性資料（例如--cryptkey）。

使用方法取決於您的操作系統和ASF版本。

Generic（通用）:

dotnet ArchiSteamFarm.dll --參數 -- 另一個參數

Windows:

.\ArchiSteamFarm.exe --參數--另一個參數

Linux/macOS:

./ArchiSteamFarm --參數--另一個參數

命令列參數也可用於通用助手腳本中，例如ArchiSteamFarm.cmd或ArchiSteamFarm.sh。 In addition to that, you can also use ASF_ARGS environment property, like stated in our management and docker sections.

若您的參數包含空格，請務必使用引號將其括住。 兩個錯誤示例：

./ArchiSteamFarm --path /home/archi/My Downloads/ASF # Bad!
./ArchiSteamFarm --path=/home/archi/My Downloads/ASF # Bad!

兩個正確示例：

./ArchiSteamFarm --path "/home/archi/My Downloads/ASF" # OK
./ArchiSteamFarm "--path=/home/archi/My Downloads/ASF" # OK

--cryptkey<key>或--cryptkey=<key>──將使用值為<key>的自訂密鑰啟動 ASF。 此選項會影響**安全性**，並將導致ASF使用您的自訂密鑰 <key>，而不是硬編碼在程式中的預設值。 Since this property affects default encryption key (for encrypting purposes) as well as salt (for hashing purposes), keep in mind that everything encrypted/hashed with this key will require it to be passed on each ASF run.

There is no requirement on <key> length or characters, but for security reasons we recommend to pick long enough passphrase made out of e.g. random 32 characters, for example by using tr -dc A-Za-z0-9 < /dev/urandom | head -c 32; echo command on Linux.

It's nice to mention that there are also two other ways to provide this detail: --cryptkey-file and --input-cryptkey.

Due to the nature of this property, it's also possible to set cryptkey by declaring ASF_CRYPTKEY environment variable, which may be more appropriate for people that would want to avoid sensitive details in the process arguments.

--cryptkey-file <path> or --cryptkey-file=<path> - will start ASF with custom cryptographic key read from <path> file. This serves the same purpose as --cryptkey <key> explained above, only the mechanism differs, as this property will read <key> from provided <path> instead. If you're using this together with --path, consider the fact that relative path will be different depending on the order of arguments, i.e. whether you switch --path before or after --cryptkey-file.

Due to the nature of this property, it's also possible to set cryptkey file by declaring ASF_CRYPTKEY_FILE environment variable, which may be more appropriate for people that would want to avoid sensitive details in the process arguments.

--ignore-unsupported-environment - will cause ASF to ignore problems related to running in unsupported environment, which normally is signalized with an error and a forced exit. Unsupported environment includes for example running win-x64 OS-specific build on linux-x64. While this flag will allow ASF to attempt running in such scenarios, be advised that we do not support those officially and you're forcing ASF to do it entirely at your own risk. It's important to point out that all of the unsupported environment scenarios can be corrected. We strongly recommend to fix the outstanding problems instead of declaring this argument.

--input-cryptkey - will make ASF ask about the --cryptkey during startup. This option might be useful for you if instead of providing cryptkey, whether in environment variables or a file, you'd prefer to not have it saved anywhere and instead input it manually on each ASF run.

--minimized - will make ASF console window minimize shortly after start. Useful mainly in auto-start scenarios, but can also be used outside of those. This option requires appropriate environment support - it may not work properly in all possible scenarios.

--network-group <group> or --network-group=<group> - will cause ASF to init its limiters with a custom network group of <group> value. This option affects running ASF in multiple instances by signalizing that given instance is dependent only on instances sharing the same network group, and independent of the rest. Typically you want to use this property only if you're routing ASF requests through custom mechanism (e.g. different IP addresses) and you want to set networking groups yourself, without relying on ASF to do it automatically (which currently includes taking into account WebProxy only). Keep in mind that when using a custom network group, this is unique identifier within the local machine, and ASF will not take into account any other details, such as WebProxy value, allowing you to e.g. start two instances with different WebProxy values which are still dependent on each other.

Due to the nature of this property, it's also possible to set the value by declaring ASF_NETWORK_GROUP environment variable, which may be more appropriate for people that would want to avoid sensitive details in the process arguments.

--no-config-migrate - by default ASF will automatically migrate your config files to latest syntax. Migration includes conversion of deprecated properties into latest ones, removing properties with default values (as they have no effect), as well as cleaning up the file in general (correcting indentation and likewise). This is almost always a good idea, but you might have a particular situation where you'd prefer ASF to never overwrite the config files automatically. For example, you might want to chmod 400 your config files (read permission for the owner only) or put chattr +i over them, in result denying write access for everyone, e.g. as a security measure. Usually we recommend to keep the config migration enabled, but if you have a particular reason for disabling it and would instead prefer ASF to not do that, you can use this switch for achieving that purpose. Keep in mind however, that providing correct settings to ASF will become from now on your new responsibility, especially in regards to deprecations and refactors of properties in future ASF versions.

--no-config-watch - by default ASF sets up a FileSystemWatcher over your config directory in order to listen for events related to file changes, so it can interactively adapt to them. For example, this includes stopping bots on config deletion, restarting bot on config being changed, or loading keys into BGR once you drop them into the config directory. This switch allows you to disable such behaviour, which will cause ASF to completely ignore all the changes in config directory, requiring from you to do such actions manually, if deemed appropriate (which usually means restarting the process). We recommend to keep the config events enabled, but if you have a particular reason for disabling them and would instead prefer ASF to not do that, you can use this switch for achieving that purpose.

--no-restart - by default ASF follows AutoRestart config property, which you can use for specifying whether restart is allowed when required. Some solutions that we provide take charge of process management and are explicitly incompatible with auto-restart function of ASF, such as running ASF in docker or systemd, as they require process to exit only, since it's their responsibility to restart it afterwards, if deemed appropriate. Since arbitrary config edit is unwanted from user experience, this switch simply overrides your AutoRestart config property by explicitly initializing it to false, even if you've specified otherwise in the config. Thanks to that, ASF can be informed in advance about running in such environment, without a requirement of providing a compatible AutoRestart: false config file.

In addition to the above, --no-restart, in contrary to AutoRestart: false, will also forbid you from using restart command or otherwise issuing ASF process to restart, since the switch explicitly states it's not compatible with such setup, while AutoRestart property only specifies default behaviour.

Normally you can (and should) control the behaviour explained here in the config file, although if you're running ASF inside a custom script or other similar environment, you may also want to make use of this switch, that will forbid ASF from restarting itself.

--no-steam-parental-generation - by default ASF will automatically attempt to generate Steam parental PINs, as described in SteamParentalCode configuration property. However, since that might require excessive amount of OS resources, this switch allows you to disable that behaviour, which will result in ASF skipping auto-generation and go straight to asking user for PIN instead, which is what would normally happen only if the auto-generation has failed. Usually we recommend to keep the generation enabled, but if you have a particular reason for disabling it and would instead prefer ASF to not do that, you can use this switch for achieving that purpose.

--path <path>或--path=<path>──ASF在啟動時始終會導航至自身所在的目錄。 By specifying this argument, ASF will navigate to given directory after initialization, which allows you to use custom path for various application parts (including config, logs, plugins and www directories, as well as NLog.config file), without a need of duplicating binary in the same place. 如果您想將二進位檔案和實際配置檔案分開，這可能會非常有用，類似Linux 打包機制——這樣您就可以在多個設置中共用一個（最新的）二進位檔案。 此路徑既可以是基於當前 ASF 二進位檔案所在位置的相對路徑，也可以是絕對路徑。 Keep in mind that this command points to new "ASF home" - the directory that has the same structure as original ASF, with config directory inside, see below example for explanation.

Due to the nature of this property, it's also possible to set expected path by declaring ASF_PATH environment variable, which may be more appropriate for people that would want to avoid sensitive details in the process arguments.

If you're considering using this command-line argument for running multiple instances of ASF, we recommend reading our management page on this manner.

範例:

dotnet /opt/ASF/ArchiSteamFarm.dll --path /opt/TargetDirectory # Absolute path
dotnet /opt/ASF/ArchiSteamFarm.dll --path ../TargetDirectory # Relative path works as well
ASF_PATH=/opt/TargetDirectory dotnet /opt/ASF/ArchiSteamFarm.dll # Same as env variable

├── 📁 /opt
│     ├── 📁 ASF
│     │     ├── ⚙️ ArchiSteamFarm.dll
│     │     └── ...
│     └── 📁 TargetDirectory
│           ├── 📁 config
│           ├── 📁 logs (generated)
│           ├── 📁 plugins (optional)
│           ├── 📁 www (optional)
│           ├── 📄 log.txt (generated)
│           └── 📄 NLog.config (optional)
└── ...

--service - this switch is mainly used by our systemd service and forces Headless of true. Unless you have a particular need, you should instead configure Headless property directly in your config. This switch is here so our systemd service won't need to touch your global config in order to adapt it to its own environment. Of course, if you have a similar need then you may also make use of this switch (otherwise you're better with global config property).

--system-required──聲明此開關將導致 ASF 嘗試通知操作系統「此進程要求系統在其存留期內處於啟動狀態並正常運行」。 This can be proven especially useful when farming on your PC or laptop during night, as ASF will be able to keep your system awake while it's running. This feature is currently supported on Linux and Windows.

On Linux, you'll need properly working dbus with login daemon that supports Inhibit() function. Two most popular daemons, systemd as well as elogind, are confirmed to support that. You can use commands such as systemd-inhibit --list or elogind-inhibit --list for confirmation whether ASF is properly holding the inhibit lock after starting. Majority of desktop environments (such as Gnome or KDE) should work out of the box, as they already depend on this functionality for their own needs - other DEs may require installation of dbus, if for some reason it's not on your machine yet.

No special requirements on Windows, it should work out of the box.