Official Manual · Settings Reference

Shadowrocket
Settings Guide

Check settings by scope, record the original values, then verify the result with the same request.

For your first setup, start with the Quick Start Guide to import your existing details and verify the connection. This page is for reference after you connect: it explains how settings interact, suggests baseline practices and shows how to roll back changes. The steps below use iPhone / iPad as a reference; options may be located elsewhere on other platforms, and system permission steps may not apply directly.

Shadowrocket is available for iPhone / iPad / Mac / Windows / Android / Linux. The iPhone / iPad app is a one-time $2.99 purchase on the App Store. The Mac (Apple Silicon and Intel) / Windows / Android / Linux versions are free. For downloads and installation, see the Downloads for All Platforms page. For iPhone / iPad, check the App Store page for system requirements.

01 / SETTINGS

Settings Scope and Default-Value Checks

Separate server details, configuration files and app settings

Shadowrocket’s behavior is not controlled by a single switch. Home stores servers and details imported through Subscribe; Config manages rules and related configuration; Settings provides app-level behavior and tools. VPN and network permissions, along with the current Wi-Fi or cellular network, add another layer of conditions. When troubleshooting, first identify which layer is involved: check server details for an incorrect address, rules for unexpected domain routing, and On Demand for automatic reconnections. Not every issue is a Settings issue.

The same setting can also appear in the app and in a configuration file. For example, DNS may be affected by the active Config as well as the system network environment. A value shown in Settings does not necessarily mean the active connection uses that value alone. First record the current Global Routing mode, selected server and active Config. Then check whether the configuration file defines related parameters, and finally review app settings. Comparisons are meaningful only after you identify which configuration sources are in effect.

“Default” does not necessarily mean the value currently shown on screen

This guide does not treat the initial state on one device as a universal factory default. App updates, migrations, subscription-provided settings and restored backups can all change current values; option names and available choices may also vary by platform. Where there is no consistent value to verify, the recommended baseline is to keep the current working setup—not to invent a default number. If the interface offers an explicit reset option, back up first and note the differences before and after resetting.

To establish a baseline, keep a local record of the network type, Global Routing mode, active Config name, server label and the setting you plan to change. Do not include passwords, full subscription URLs or PrivateKey values in records you intend to share. Before testing, choose a destination that already works and use that same destination after the change. Do not change the website, network and server at the same time; otherwise, even an improved result will not tell you which change made the difference.

ScopeWhat it coversRecommended baseline
HomeServers, Subscribe and current selectionKeep one verified working entry
ConfigRules, DNS and configuration referencesEdit a copy; do not overwrite a working file
SettingsTesting, logs, sync and local behaviorChange only settings related to the issue
System settingsVPN, network permissions and background conditionsGrant permissions when prompted

Make every change reversible

Before editing, export or copy the current Config. If server details are involved, make sure you can restore them from the original source. A screenshot of the Home list is not enough; it usually does not include all parameters. Then change one variable at a time. Disconnect and reconnect as needed, and confirm the new setting is active. A common reason for a change to have no effect is editing a file that is not selected. Checking the filename alone is not enough; verify the file contents too.

If a change causes a problem, restore that one value before trying anything else. Test again; a clear link is established only if the issue disappears when the setting is restored. If you cannot find the relevant option, do not guess at a similarly named setting as a substitute. Check the Glossary first. When maintaining a shared device, also explain which connections the change may affect so others do not change the active server or configuration during testing.

02 / ROUTING

Global Routing and Config Rule Priority

What Config, Proxy and Direct change

Global Routing determines how requests are handled overall. Config uses the active Config to route requests according to its rules. Proxy is useful for briefly checking whether a request can use the selected proxy route. Direct lets you compare the result on the current network without a proxy. These are not performance levels, and “more global” does not mean “more complete.” Config is usually the right choice for everyday rule-based routing. Switch to the other modes temporarily for comparison while troubleshooting, then restore the original mode.

Distinguish the Config mode from the Config tab at the bottom of the app: the mode determines whether rules are used, while the tab manages configuration files. Even if you have edited and selected a configuration file, you cannot tell whether a domain rule matched correctly if Global Routing is set to Proxy. Conversely, an issue in Config mode does not prove the server is unavailable: an earlier DIRECT or REJECT rule may handle the request before it reaches the intended proxy route.

Description / interface labelWhat to checkWhat it does not prove
ConfigVerify rule matching in the active configurationThat every request goes through the selected server
ProxyCheck whether the selected proxy route is availableThat every rule in Config is correct
DirectCompare direct access on the current networkThat every destination is suitable for direct access

Read rules in order; do not infer their purpose from their names

Rules are generally evaluated in the order they appear in the configuration. The first matching rule determines how a request is handled. DOMAIN matches an exact domain; DOMAIN-SUFFIX matches a domain suffix; IP-CIDR and IP-CIDR6 match address ranges; GEOIP classifies by IP geolocation database; FINAL handles requests not matched earlier. Do not include a scheme in a DOMAIN-SUFFIX value, or add a URL path or query parameters. To distinguish a root domain from its subdomains, decide the intended match scope before choosing a rule keyword.

This snippet illustrates syntax and order only: the first domain suffix uses PROXY, matching GEOIP requests then use DIRECT, and requests not matched earlier use the PROXY specified by FINAL. It is not a complete configuration for general use and contains no server credentials. GEOIP rules may depend on the resolved address and database results; they are not a substitute for explicit domain-based rules. If a destination matches an earlier domain rule and a later address rule, the order matters—the later policy does not automatically take precedence.

[Rule]
DOMAIN-SUFFIX,example.com,PROXY
GEOIP,CN,DIRECT
FINAL,PROXY

Verify rule matching with a real request

To check a domain, search the active configuration for the exact domain, parent suffixes and related rule references. Look for an earlier rule that may match first. Then eliminate testing variables: close the destination page and start a new request; disconnect and reconnect if needed to avoid an existing connection continuing to carry traffic. A browser page may request content from several domains. If the page loads but images fail, that does not mean the whole site matched a single rule. Find the actual failing hostname in the logs and check each one separately.

Start by adding a narrowly scoped test rule to a copy of the configuration. Once you confirm that it matches, decide whether to merge it into the working file. Do not casually change FINAL to another policy to fix one domain; that affects every unmatched request. Check the intended use of the existing configuration to choose its default policy—there is no universally correct value. If the site works in Proxy mode but fails in Config, compare matching rules, DNS and the policy target first. If it fails in both, check the server, authentication details and underlying network.

After updating rules, check that remote resources were retrieved successfully and the configuration was reloaded. Updating the server list through Subscribe is not the same as updating rules in Config; a successful list refresh does not prove that the rules are active. To establish a basic connection first and revisit rule matching later, follow the Global Routing steps in the tutorial. For specialized keywords, consult the Glossary rather than inferring rule behavior from a Chinese note.

03 / DNS

DNS Resolution, Caching and Request Paths

DNS resolves names; it does not control all traffic

DNS translates domain names into usable addresses and is one part of the connection path. A successful resolution only confirms that an answer was returned; it does not prove that the destination port is reachable, the TLS handshake succeeded or the webpage loaded completely. Conversely, a page that will not load is not necessarily a DNS problem: server authentication may have failed, a rule may have sent traffic the wrong way, or the destination service may be down. Identify the stage where the problem occurs before changing DNS servers repeatedly.

First check whether the active Config explicitly defines DNS, then review the corresponding item in Settings. If you have no specific need, keep the current working setup rather than treating a public resolver as the right answer for every network. System-provided resolvers, manually specified resolvers and encrypted DNS each have their own reachability requirements. The resolver itself must also be accessible through the current network path. Adding an unreachable address can turn an issue affecting one site into a problem with multiple domains.

Recording the default state and recommended baseline

Use the current installation, active Config and actual system network settings to determine the DNS default. Record not just the address but whether it comes from the system, app settings or a configuration file. If you specify multiple addresses, preserve their full order and related options. In the example below, system shows the configuration syntax for using system DNS; it does not mean every device ships with that line in its default file. Do not paste this snippet over an existing complete file, especially if that would remove its servers or rules.

[General]
dns-server = system

[Rule]
DOMAIN-SUFFIX,example.com,PROXY
FINAL,DIRECT

A good troubleshooting baseline is to keep one resolver setup you have confirmed is reachable and avoid changing several DNS options at once. If you switch from system DNS to a manual resolver, retest with the same domain, network and server, and check whether the returned address changes. A different address is not automatically an error; the same service may return different results based on location or load. Compare the DNS answer, the subsequent connection and the actual access result to judge whether the change helped.

Caching, IPv6 and local network names

DNS responses may be cached at different levels by the app, system and tools used to access a site. If you still see an old address after changing a setting, the configuration may be working; an existing connection may not have closed, or the cache may still be valid. If the interface offers a cache-clearing option, record the old result, clear the cache, reconnect and make a new request. If there is no such option, do not guess at a hidden procedure. Refreshing a page may reuse the existing connection, so check the logs to confirm that a fresh resolution and connection occurred.

For a destination that returns both IPv4 and IPv6 addresses, one reachable address does not prove that everything is working. The network may provide an IPv6 address but fail to connect reliably, or may restrict only one address family. Record the actual connection address and its type, then compare under the same conditions. Do not start by permanently disabling an address family. For local hostnames, router management addresses or internal company domains, check whether they depend on a resolver available only on the current network.

A local hostname may not resolve correctly after switching to an external resolver; that does not mean the remote server details are invalid. First test the name directly on the original network, then check whether a rule sends internal traffic along an unsuitable path. Captive portals may also rely on system DNS and redirection behavior. Avoid major DNS changes before completing network authentication. Once connected to the network, test again with Shadowrocket to distinguish environmental issues from configuration problems.

04 / ON DEMAND

On Demand and the Connection Lifecycle

Treat automatic triggers and rule-based routing separately

On Demand triggers connection behavior based on configured conditions; it manages the connection lifecycle. Global Routing determines how requests are handled once a connection is established. They are separate features. Enabling On Demand will not fix incorrect server details or create complete routing rules for you. First confirm that the server, Config and access to your destination all work with a manual connection. Otherwise, automation will simply repeat the same failed attempt and make troubleshooting harder.

Check the current device interface and system VPN configuration to see whether On Demand is enabled, especially after restoring a backup or changing devices. During troubleshooting, use a manual connection temporarily. Record the original On Demand switch and conditions before testing a network change. Once the manual connection is stable, restore automatic triggers one at a time as needed. Do not enable every condition without testing just to make the connection appear “always on.” The more conditions you add, the more important it is to understand how they match.

Why a network change can trigger a reconnect

When switching from Wi-Fi to cellular, the local address, routes and available resolvers may all change, so existing connections may need to be re-established. A brief status change does not necessarily mean the configuration is broken. Wait for the connection process to finish, then check an actual request. If every network switch fails but the connection works consistently on the original network, compare whether each network can reach the server address and port, and whether their DNS paths differ. Do not focus only on the automatic trigger switch. On Demand can initiate an attempt, but it cannot change underlying network conditions.

If your settings use network name or type as a trigger condition, first confirm that the current connection actually meets it. Two Wi-Fi networks with similar names may fall into different rule scopes. Whether the system provides network details to the app also depends on permissions and system behavior. Test one clearly defined scenario—for example, observe one connection on the current Wi-Fi, then switch to another known network. At each step, record whether a connection was triggered, whether it succeeded and whether retries continued after failure.

  1. Turn off automatic triggers first, then confirm that a manual connection and actual access both work.
  2. Record the existing conditions and keep only the On Demand condition you need to test.
  3. Switch networks once, wait for the connection state to settle, then check a new request.
  4. Restore the original values after testing so temporary conditions do not become everyday settings.

Diagnosing a reconnect after manual disconnection

If the connection comes back soon after you manually disconnect, first check whether On Demand is still enabled and whether the system VPN settings retain a corresponding on-demand behavior. Then check whether another authorized automation is calling the connection action. Do not repeatedly toggle the switch to “hold” the connection off; this creates multiple connect and disconnect events in a short time, making logs harder to interpret. Stop the automatic trigger first, then perform one deliberate manual action for a clearer result.

A system VPN indicator can provide a clue about connection status, but it cannot prove which rule handled a destination’s traffic—or that existing service details are still valid. Check Shadowrocket’s connection status and make an actual request as well. Device lock, temporary network outages and system resource management can all affect connection behavior; do not assume the same session will remain active in every situation. For long transfers, observe how the connection recovers after a network switch, not just whether a page opens initially.

If automatic connections keep failing on Wi-Fi that requires network sign-in, turn off On Demand, complete the system’s network authentication, then test a manual connection. Remember to restore the original settings after leaving that network. If the issue occurs only on a particular network, record its type and the order of events; you do not need to share its full name or your location. For everyday use, keep only the minimum set of clearly understood triggers. After changing devices or editing Config, check that the conditions still match your intended use.

05 / CONNECTIVITY TEST

Ping, Connectivity Test and Timeout Diagnosis

Similar test names may measure different things

In everyday usage, Ping is often used as a general term for a latency test, but the app’s specific tests may involve different processes. Network-layer probing, a connection test to a server port and a test that accesses a destination through a proxy are not interchangeable. Before using Connectivity Test, check the method and target described in the current interface. A time value beside an entry does not prove authentication, DNS, rule matching and webpage loading have all been verified. The key is which layer the test actually covers.

Use the current interface to determine the default test method, target and timeout; this guide does not prescribe unverified factory values. Keep the existing working values unless the target is unreachable, responds too slowly or is unsuitable for the current network. If you need to enter a test address, choose one you can verify that responds reliably and does not require sign-in, and check whether the test uses a proxy. Do not use a link containing private parameters, access tokens or temporary authorization details as a public test address.

Assess latency, throughput and service availability separately

A short response time reflects the result for a particular target and test method at that moment; it does not mean large files will transfer faster. Throughput also depends on link bandwidth, congestion, server limits and destination load. Page loading may involve multiple domains and rounds of requests. When choosing among servers you already have, first confirm that the destination service works, then compare results across repeated tests under the same conditions. Do not switch based only on the lowest value from one test, or treat a test result as a long-term guarantee.

Observed resultCheck firstNext verification
Test times out, but actual access worksTest target, method and timeoutKeep the actual request unchanged and compare with another test method
Test responds, but webpage failsAuthentication, DNS, rules and destination serviceCheck the logs for the specific failing connection stage
Fails only on one networkServer port and network pathRepeat the same test on another known network
Intermittent variationNetwork changes, concurrent transfers and destination loadRepeat at different times under the same test conditions

A timeout does not mean the server has failed. The test target may not respond as expected, the current network may be unable to reach the server port, or the timeout may be shorter than the time needed to complete the test. First check whether other access is failing too, then determine the scope of the issue. If only one test method fails, check how that test works. If all requests fail, check the server address, port, password or key, and confirm that the device has actually established a connection.

Build a reproducible comparison test

Before running tests in batches, stop large file transfers, keep the device on the same network if possible and test a small number of existing entries. Concurrent tests consume connection resources, and results may vary with device and network load; do not rank data from different batches as if they were directly comparable. Test the same entry several times and note whether each run completes. If only the first run is slow, connection setup, resolution or caching may be involved. Check the logs before changing protocol parameters.

To compare two settings, record the original method, target and result, then change one value only. Do not update Subscribe, change Global Routing and switch DNS during the same test. When finished, repeat an actual task—for example, reopen the same page and wait for its main content to load. Use the results for everyday decisions only when the test and real access move in the same direction. If they conflict, keep the setup that completes the actual task and continue investigating the difference.

If you have multiple Subscribe entries, identical labels do not mean their connection parameters are the same. Before testing, confirm which source each entry belongs to. Updates may also create duplicates, making it seem as if you are comparing different servers when you are not. See the Managing Multiple Subscriptions article for help with grouping, sorting and duplicate entries. This section covers testing methods only; it does not evaluate or recommend any service source, or present a test list as proof of service quality.

06 / LOG

Log, Data and Step-by-Step Troubleshooting

Narrow the time window before reading errors

Log helps you inspect the connection process. Traffic records in Data can help show whether data was transferred. Neither a log nor a total traffic count alone proves that a destination was accessed successfully. Log locations and visible fields depend on the current interface. Note when you reproduce the issue, then perform the test action once. When a large amount of history is mixed in, it is easy to mistake a background request error for the cause of the current page failure. Narrow the observation window first instead of reading every line from the start.

Check the current settings to determine the default log detail level. For everyday use, keep only the detail needed for basic diagnosis; temporarily increase it only when reproducing a complex issue. If the interface has no log-level option, do not assume a switch described elsewhere exists. Detailed logs may increase storage and processing overhead, and expose more browsing context. Restore the original setting once you have identified the issue, and manage existing records according to your privacy needs. Do not collect more information than you need over the long term.

Separate resolution, connection, authentication and app response

First check whether the request appears at all. If there is no related entry, confirm that the app is connecting, that the destination is not reusing an existing session and that its traffic falls within the current handling scope. Next, check whether an address was resolved, then whether a connection to the server or destination port was established. Resolution failure and port connection failure occur at different stages and need different responses. A “timeout” alone does not prove the password is wrong; timeouts can occur at several points in the connection path.

For authentication failures, check the server details you already have: protocol, port, password or key, encryption parameters, and TLS and transport settings if required. Do not copy fields from one protocol to another. If the connection is established but fails during TLS, check the server-required name, certificate validity requirements and device time. Disabling verification is not a general fix. If the destination returns an application-level error, investigate the destination’s permissions or service status.

[Rule]
DOMAIN,example.com,PROXY
FINAL,DIRECT

Use the narrow rule above for a comparison in a copy of the configuration: route only the exact domain through PROXY and handle the rest with FINAL. It does not cover possible subdomains used by the site and is not a complete troubleshooting configuration. If the main page and static resources use different routes, the logs will show multiple hostnames; check each one separately. When finished, remove the temporary rule or restore the original file so the test FINAL does not change everyday request routing. Saving the configuration and confirming it is the active one are both necessary.

Turn logs into a reproducible issue report

A useful issue report includes the device platform, network type, Global Routing mode, purpose of the active Config, the action that failed and the comparison tests already completed. Check the version in the app and report the value you actually see; do not copy it from an old article. Preserve the original error text and context, but remove full subscription URLs, authentication fields, personal domains and network names before sharing. Removing only the password may not be enough if a link still contains an access token.

If Config works for some destinations, compare the rules matched and address types for successful and failed requests. If only one server entry fails, compare it with the original details. If every entry fails on one network but works on another, investigate the network path first. These branches help narrow down the issue. Reinstalling repeatedly, deleting all configurations or resetting multiple options at once can destroy useful evidence and make a recoverable state difficult to restore.

After troubleshooting, run a regression check: restore the log detail level and remove temporary rules, verify that previously working destinations still load, confirm that no test conditions remain in On Demand and check for unexpected repeated reconnections. Describe the result in terms of what you observed—for example, “the port connection worked after switching networks”—rather than vaguely saying “the settings fixed it.” If the evidence is incomplete, say so and collect another single-run reproduction instead of presenting a one-off success as a guaranteed fix.

07 / BACKUP

iCloud Sync, Config Backups and Migration

Sync is not a standalone historical backup

iCloud sync transfers app-supported data between eligible devices, but check the current interface and actual restore results to see what is supported. Do not assume that all servers, rules, logs and system permissions migrate together. Sync is also not a way to return to any arbitrary previous state: deletions or incorrect changes may propagate. Before extensively reorganizing Config, deleting entries in bulk or changing devices, make a separate export backup and confirm that you can open the file.

Check sync status separately in the app, system iCloud permissions and account settings. Signing in to an account does not prove that Shadowrocket has synced. Keep the current working device unchanged at first, check for the expected content on another device, then decide whether to adjust the options. If a feature is available only on certain platforms, use the entry point provided on that platform. When moving between platforms, do not assume a cloud-sync feature on one system is a universal file-transfer method.

Back up rules, server details and settings separately

A Config file primarily contains configuration, but it may only reference servers or remote resources rather than include all their details. Before exporting, inspect the file and determine whether it contains rules, DNS parameters, resource references, authentication details or some combination. Saving a file with a name is not proof of a complete backup; saving a Subscribe URL alone does not guarantee that its contents will still be available later. Keep original details somewhere safe, and do not put real credentials in public notes or shared folders.

Prepare for migration in three parts: first, save a copy of the working Config; second, keep the information needed to recreate your existing server entries; third, record app-level settings such as Global Routing, On Demand, test methods and the purpose of local ports. Store exported files somewhere access is controlled. You can label files by purpose, but do not include passwords or subscription tokens in their names. Do not casually send sensitive backups to others or rely on public cloud links for long-term storage.

  1. On the old device, confirm that the current connection works, then export Config and inspect the file contents.
  2. Separately verify that you have the original information needed to restore existing servers and Subscribe entries.
  3. After importing on the new device, select the Config, Global Routing mode and current server again.
  4. Test actual access and automatic connections before cleaning up the old device and duplicate files.

Resolving Sync Conflicts and Checking Restores on a New Device

Editing the same configuration on two devices at once may cause overwrites or hard-to-distinguish copies. Before resolving a conflict, keep both versions and compare their rules, resource references and key settings. Then choose which one to use as the starting point. Matching filenames do not prove that contents match. Do not repeatedly delete and recreate files before sync has finished. Choose one device for the current editing session and pause changes on the other until you have confirmed the result; this makes it easier to trace where changes came from.

After restoring, connect manually before enabling On Demand. System VPN authorization usually needs to be confirmed again on a new device; check local network access and other system permissions when prompted. Then confirm that Global Routing suits its original purpose, the correct file is selected in Config and the server points to a valid entry. If rules reference remote resources, check that those resources can be retrieved. Seeing a file in the list does not mean all of its external dependencies are ready.

If the old device works but the new one does not, compare them item by item rather than immediately assuming sync failed. The devices may be on different networks, use different DNS conditions or have entries with the same name but different parameters. Put them in comparable environments first, then check configuration contents and current selections. Also consider whether the existing service limits simultaneous connections. That is a condition of your existing service; changing the client’s display name will not resolve it. Do not share account credentials for testing.

For the full migration sequence, see the Config Backup and Device Migration article. Before cleaning up old data, restart the app and verify access at least once, then confirm that the backup is still available independently. If installation on the new device is not complete, follow the downloads page for app acquisition and system prompts before restoring configuration. App installation, data restoration and connection verification are three separate stages. Completing them separately helps distinguish installation, sync and network issues.

08 / LOCAL ACCESS

Proxy Ports, Widget and Everyday Maintenance

A local Proxy port is not a server port

A local Proxy port lets eligible programs connect to the client using a specified protocol. The server port in Home is a parameter for connecting to a remote service. They may happen to use the same number, but there is no requirement that they match. If a program asks for an HTTP or SOCKS5 proxy, first check whether Shadowrocket on your platform offers the relevant listener and whether it is enabled. Do not enter the remote server address and port as the local proxy unless that is how your existing service is meant to be accessed.

Use the current Settings interface to check the port value, listening address and default switch; this guide does not prescribe a number that applies to every device. As a baseline, do not enable LAN sharing when no program needs an explicit proxy. If only a program on the same device needs it, limit access to the device where possible. Before changing the listening port, record its original value and make sure another program is not using the new port. Update any programs that use it too. If only one end is changed, the other will still point to the old port and may be refused or time out.

ItemMeaningWhat to check
Listening addressWhich local network interfaces accept connectionsLocal device and LAN access have different scopes
Local portWhere the client receives proxy requestsThe port entered in the calling program must match
Proxy protocolThe protocol agreed on by the calling program and local listenerHTTP and SOCKS5 are not interchangeable
Server portConnection details for an existing remote serviceUse the server details you already have

Limits and Verification for LAN Sharing

Allowing LAN devices to connect expands who can access the listener, so enable it only when needed and on a controlled network. First confirm that the connecting device and the device running Shadowrocket can reach each other, then check the system firewall, listening address and port. Being on the same Wi-Fi network does not guarantee that devices can communicate; guest network isolation may block them. Do not weaken all system protections just to test one program, or expose the local listener to external networks unnecessarily.

When entering a proxy address, “localhost” means different things to a program on the same device and one on another device. A local loopback address in the calling program points to the device running that program; it will not automatically point to a different device running Shadowrocket. For access across devices, use a reachable LAN address and account for the possibility that it may change after reconnecting to the network. When testing is complete, disable sharing if it is no longer needed, restore the original listening scope and remove any leftover proxy settings from the calling program.

Widget is a shortcut, not a connection test

Widget provides a convenient way to view or trigger actions. Available actions depend on the components offered by the current system and app. Before adding one, confirm in the app that the connection works, then add the Widget using the system’s widget controls. System scheduling can delay Widget updates, so its text or icon at any moment is not a real-time network monitor. If the display disagrees with actual access, open Shadowrocket to check the connection and server, then make a request. Do not repeatedly toggle the connection through the Widget.

Whether to add a Widget depends on the system layout and user choice; there is no universal setting in a Config file. Keep only shortcuts with a clear purpose, and be mindful of server labels shown on the Lock Screen or shared screens. After changing devices or restoring a backup, you may need to set up the Widget and permissions again. Even if the Widget reappears, check Config, Global Routing and the current server. A Widget can trigger an action without restoring the original server details or system permissions.

For routine maintenance, follow a consistent order: confirm that your existing service details are valid, check the active configuration and rule references, review DNS, automatic triggers and test settings, then check local ports and Widget. Change only the layer related to the current issue, and keep a working copy and rollback notes. To get the app again, visit the Downloads for All Platforms page. If you have just completed your first setup, return to the Quick Start Guide to verify the connection. A stable baseline makes it easier to identify later changes than continually adding more settings.