Official Guide · Browse by Stage

Shadowrocket
Complete User Guide

Get the app and grant permissions → Import and route traffic → Verify and maintain. Instructions use the iPhone/iPad interface as an example.

For your first connection, start with the Quick Start Tutorial to follow a single setup path. This guide explains the prerequisites, settings, and troubleshooting steps for each stage, so you can return to the relevant section when needed.

  • 8 stages
  • An existing subscription or server details
  • Interface labels remain in English

01 / Get the App & Install

Get Shadowrocket for your device

App access and connection details are separate

Shadowrocket is a closed-source commercial application and a cross-platform rule-based proxy client. On iPhone and iPad, it costs $2.99 as a one-time App Store purchase; the Mac, Windows, Android, and Linux versions are free. The App Store manages purchases and updates for iPhone and iPad, while installers for the other platforms are available from this site's download page. The app does not include access to servers. After installation, you still need your own connection details.

Start at the download page for all platforms to choose the right version for your device; don't judge a file's source by its name in search results. Desktop and Android installers are provided on this site under /dl/shadowrocket/, and version details are read dynamically from the download page. If a download is interrupted, reopen that page. Don't mix files from other sources with this site's downloads, or assume an installer is right for your system based on a similar-looking icon.

iPhone/iPad: Check the App Store listing

On the App Store product page, check the product name, official icon, developer Shadow Launch Technology Limited, and app ID 932747118. Refer to the App Store listing for system requirements. Before paying, confirm the Apple Account and store region you're using; the store shows the final currency. If the app doesn't appear in search, check your account region, age restrictions, and account status before trying again. Don't make repeated purchases.

After the App Store download finishes, the app should appear on your device. When switching devices or reinstalling, first look in Purchased under the original account. Purchase history and connection settings are separate: restoring the app does not automatically restore subscriptions, servers, or rules. For the store link and purchase checks, see How to Buy Shadowrocket on the App Store. The final chapter of this guide covers configuration recovery.

Mac, Windows, Android, and Linux: Choose an installer and handle system warnings

On Mac, open About This Mac and choose the DMG for Apple silicon or Intel based on the chip information; don't guess based on the computer's appearance or purchase year. On Windows, use the 64-bit installer; on Android, use the arm64 package. On Linux, choose deb, rpm, or AppImage for your distribution. Multiple formats for one platform reflect different installation methods; you don't need to install every file.

Desktop and Android installers are unsigned, so the system may block them the first time you open them. Before proceeding, confirm the file came from this site's download page and check which file the warning refers to. In Windows SmartScreen, select More info, then Run anyway. If your device is managed by an organization and that option is unavailable, contact your administrator; don't change organizational policies just to install the app.

System warningWhere to handle it
macOS can't verify the developerAfter trying to open the app, go to System Settings > Privacy & Security and check for Open Anyway.
macOS says the app is damagedAfter confirming the source and path, use the command below to remove the quarantine attribute, then open the app again.
Android blocks the installationTemporarily allow Install unknown apps for the browser or file manager used to download the file.
Linux AppImage won't runMake the downloaded file executable. Install deb and rpm packages with your distribution's package manager.
xattr -cr /Applications/Shadowrocket.app

The macOS command above applies only when the app is in the directory shown. It recursively removes extended attributes from that app's directory; don't change the target to your entire Downloads or system directory. After installing on Android, you can turn off the installation permission you enabled for the browser. For a Linux AppImage, run chmod +x on the downloaded file itself. Don't grant execute permission in bulk to files you don't recognize.

Installation is complete when the app opens normally and you can access the available screens, such as Home, Config, and Settings. You don't need a server ready to connect at this stage. If the system still blocks the app, review the installation and system warning guide. Resolve app launch issues before setting up network access to avoid confusing system restrictions, missing details, and unreachable servers.

02 / First Launch & Permissions

Check system permissions before connecting

What VPN configuration permission does

The steps below use iPhone and iPad as examples. When you first open Shadowrocket, locate Home, Config, Data, and Settings; there's no need to change every switch right away. The first time you connect, the system asks permission to add a VPN configuration and may require your device passcode or biometric authentication. This permission lets the system create a network extension; it doesn't generate a server address or fill in subscription credentials for you.

Check the system prompt after you tap Connect yourself. Confirm that it refers to Shadowrocket, then grant permission. If you previously denied it, return to the app and start the connection again, then check the VPN configuration status in system settings. Having a configuration installed doesn't mean a connection is active. Likewise, seeing a server entry in the app doesn't mean the system has allowed a network tunnel. Check these separately.

If the permission prompt never appears, stop and check whether a related configuration is already installed, then see if device management restrictions are enabled. Work- or school-managed devices may block VPN creation; follow the administrator's policies in that case. Don't repeatedly delete network configurations used for other purposes, and don't treat an unfamiliar management profile as an ordinary connection entry, as this could affect your existing setup.

Grant permissions only when needed

When using Scan QR Code, camera access is only for scanning. To scan a code from your photo library, select the photos you need to access as prompted by the system. You don't need to enable every media permission before entering server details manually or pasting a subscription. If you need to grant permission again after denying it, find the app in system settings and check the relevant permission. Then reopen the scanning screen instead of retrying from an old error page.

When importing from the clipboard, the system may ask whether to allow pasting. Check that the clipboard contains the information you intend to import before proceeding. Subscription URLs often contain personal access credentials; don't paste them into search boxes, public forms, or group chats to “test” them. After importing, you can copy some ordinary text to replace the clipboard contents and avoid pasting the URL by mistake later.

If the app requests local network access, consider whether you need to reach devices on your LAN. Accessing a router, printer, or internal service is different from connecting to an internet server. Even with local network access enabled, routing rules or LAN restrictions may still affect connectivity. Permissions determine whether the app can perform an action; rules determine which route traffic takes. Neither replaces the other.

Set up a repeatable baseline

For the first test, use a network that you know can access the internet. Before connecting Shadowrocket, open a regular webpage once. This helps rule out obvious issues with the device clock, network sign-in, or basic name resolution. If a hotel, guest network, or public hotspot requires a web sign-in, complete it first; an in-app timeout may simply mean the network hasn't granted internet access yet.

Also check whether another connection is already using the system VPN. Network extensions on iPhone and iPad can interfere with each other, and a later connection may replace an earlier one. During testing, note which services are enabled instead of guessing from the status bar icon which app is handling traffic. If you need to temporarily disable another connection, restore it afterward for its usual purpose; you don't need to delete its configuration.

Before adding complex rules, keep a brief note of the environment: network type, system proxy or VPN status, whether you completed a web sign-in, and the exact error message you first saw. Record symptoms only; don't include passwords or full subscription URLs. This baseline helps you tell whether a problem followed a new configuration change or a change in the network itself.

At the end of this stage, the app should open, necessary permissions should be confirmed, and the basic network should work. You should also understand the difference between connection permission and server details. Next, add a service you already have. To get connected quickly for the first time, follow the Quick Start Tutorial; to understand each field, continue to the next chapter.

03 / Servers & Subscriptions

Import existing servers and Subscribe

Identify the type of connection details you have

Before you start, identify whether you have individual server settings, a single share link, a QR code, or a subscription URL containing multiple entries. They may all look like text, but each requires a different import method. A subscription fetches and updates a server list in bulk. Config may also contain rules and other settings; importing a subscription URL as a rules configuration won't automatically provide the routing behavior you expect.

Get connection details from a service you already use. Before importing, check that the link is complete. Watch for line breaks added by chat apps, missing characters at the end, and spaces accidentally copied around the URL. Query parameters may be essential and shouldn't be removed casually. If you can't identify a short string, ask the provider what format it uses instead of trying every import option.

Add Server: Keep related fields consistent

To use Add Server in Home, choose the correct protocol Type first, then enter the address and port. The address field usually takes a hostname or IP, not a full webpage URL with a path; enter the port separately. Choosing the wrong protocol can't be fixed just by changing the password, even if the address and port are correct. Check the protocol against your original details first, then verify the remaining fields.

FieldWhat to check
Address and portHostname or IP, the corresponding port, and any extra spaces.
Credentials and encryptionCheck the password, UUID, encryption method, and other fields for the selected protocol. Don't reuse values across different protocols.
TLSEnable status, Server Name, and certificate requirements, based on your existing service details.
TransportCheck WebSocket, Host, Path, and other fields as needed. Preserve capitalization and path symbols.

Shadowsocks, VMess, VLESS, Trojan, Hysteria2, WireGuard, HTTP, and SOCKS5 use different fields. Extra options on screen don't mean you have to fill them all in. In particular, don't copy TLS, Host, or Path values from another entry just to complete the form. A display name helps you identify an entry, but it's only a label; it doesn't change the server's location, protocol, or connection behavior.

If your details specify a domain for certificate verification, Server Name may differ from the connection address. Follow the service details. If you see a certificate error, check the device clock and whether the address and name match; don't disable certificate verification as a routine fix. If key fields are missing, get them from the source. Randomly changing security settings makes the issue harder to reproduce and diagnose.

Subscribe: Check the result after importing

Paste the URL into Subscribe or the subscription import option available in the current interface. Give it a recognizable name, then update it. Return to Home and check that entries appear and are grouped as expected. The subscription name only helps you identify the source; the URL and its access permissions determine what gets fetched. The interface may change, so use the import option shown in the current app.

If an update fails, first distinguish between “can't retrieve the URL” and “can't read the content.” For the former, check the network, URL, and access permissions. For the latter, check whether the response is in a format the client supports. If the URL returns a login page, an expired notice, or a verification page, refreshing it repeatedly won't turn it into a server list. Keep the error message and check the subscription status with your existing service provider.

When importing from a QR code, check the detected content type before saving. A QR code is just a way to encode text, not proof that the content is trustworthy. If the camera has trouble scanning it, use the photo library option only if the image is clear and from a known source. Don't upload QR codes containing credentials to public scanning sites or share them in public support requests; others may be able to read the encoded details.

A subscription update may add, remove, or rename entries. Before updating, note the currently selected server and distinguish subscription-managed entries from ones you added manually. Importing the same URL repeatedly can create confusing duplicate groups, so update an existing subscription instead of adding it again each time. After a successful import, choose an entry with complete details for connection testing. Don't edit every server's settings at once.

This stage is complete when the list contains identifiable servers, key fields match the original details, and the subscription update has a clear result. The number of entries alone doesn't indicate connection quality. For how server selection relates to test results, see Latency Tests and Server Selection. The next chapter explains how to route traffic through these entries.

04 / Global Routing

Config, Proxy, and Direct: What's the difference?

Three modes for different needs

Global Routing controls how traffic handled by the client is routed overall. The three modes are Config, Proxy, and Direct. These modes are not server protocols, nor do they indicate whether a connection is working. Switching to Proxy won't fix an incorrect password, and switching to Config won't automatically create rules that suit your needs.

ModeHow it worksCommon use
ConfigUses rules and policies in the currently enabled configuration to handle connections.Route traffic based on conditions such as domain or address.
ProxyConnections handled in this mode use the selected proxy route.Temporarily bypass routing rules.
DirectConnections handled in this mode access their destinations directly.Compare direct connection results and check whether a proxy route is being used by mistake.

For everyday routing, first check which configuration is enabled on the Config screen, then set Global Routing to Config. Downloading a file to the list doesn't necessarily enable it. Likewise, editing a file that isn't in use won't affect the current connection. Before testing, note both the file name and routing mode. This helps avoid troubleshooting the right settings in the wrong configuration.

Rules are checked from top to bottom

Rules are usually checked in order, and the first matching entry determines the policy. Put specific conditions before broader ones. DOMAIN matches a specific domain; DOMAIN-SUFFIX matches a domain and its subdomains; DOMAIN-KEYWORD checks for a keyword in the domain. IP-CIDR and IP-CIDR6 match address ranges, GEOIP matches based on IP geolocation data, and FINAL handles connections that didn't match an earlier rule.

[Rule]
DOMAIN-SUFFIX,example.com,PROXY
DOMAIN-KEYWORD,ads,REJECT
IP-CIDR,192.168.0.0/16,DIRECT
GEOIP,CN,DIRECT
FINAL,PROXY

This is a syntax example, not a complete configuration for every network. The first line sends the example domain through PROXY. The second rejects requests whose domain contains the specified keyword. The third sends the shown private IP range direct. GEOIP results depend on IP geolocation data, not a website's language. FINAL provides a fallback route and should be last in the relevant rule list.

PROXY, DIRECT, and REJECT are rule policy keywords. They are not the same as the Global Routing options, whose capitalization differs. PROXY also needs an available proxy target; DIRECT depends on the current network being able to reach the destination; REJECT actively blocks matching connections. If the configuration references a custom policy group, check that the group exists and has a valid selection.

Keep a way to roll back rule changes

Before editing, duplicate or export the working configuration. Add just one rule for a specific test target, then reconnect and check the result. Don't change a subscription, replace the entire configuration, and switch routing modes all at once; if things improve, you won't know what made the difference. Domain rules can also be affected by whether requests retain domain information, how names are resolved, and whether connections are reused. Review the logs in context.

For example, if you can't reach a LAN device, check the address range it actually uses instead of copying an example private range. Home, office, and guest networks may use different subnets. If IPv6 is in use, check which address type the connection actually uses. An incorrect address range can send some devices along an unexpected route while others still work.

A configuration may also include remote references. Before importing, understand their purpose and source. If the source changes its content, a later update may change routing behavior, so keep any exception rules you depend on. For more syntax and rule-order examples, see Rule Syntax Basics. By the end of this chapter, you should be able to identify the active configuration, overall mode, and relevant rule for a destination.

05 / Connections & Verification

Troubleshoot connection failures layer by layer

Connect first, then test access

Return to Home, select the server you want to test, check Global Routing, and connect. Grant system permission when prompted during the first connection. If permission was already granted, check that the switch stays on and that the system shows the expected status. A switch that immediately resets, a connection that keeps trying, and a webpage that fails after connecting are three different symptoms. Note where the failure occurs before checking permissions, server details, or the route to a particular destination.

Before testing, avoid actions that make results hard to interpret, such as repeatedly switching servers, refreshing a subscription, or changing several settings at once. Choose an entry with complete details and run one test on the same network before comparing another. If you change networks, destinations, and configurations with every test, results have no shared conditions. Successes and timeouts become hard to explain, and you can't build a repeatable troubleshooting process.

Connectivity Test can show reachability under a particular test method, but it doesn't fully test access to every site or app. The test target, connection protocol, and network conditions all affect the result. A test value for an entry may only mean that a particular probe succeeded. No value doesn't necessarily mean every connection will fail. Check actual access and error logs as well.

Check in four layers

  1. Confirm the device's basic network works, public hotspots are signed in, and the system clock is reasonably accurate.
  2. Check the server address, port, credentials, and transport settings, and confirm your existing service is active.
  3. Check the active Config, Global Routing mode, and policy group selection. See which route the destination actually takes.
  4. Test with a new request to a destination. Check whether the webpage, images, or other resources are failing.

For a name resolution error, first check whether the server hostname resolves in the current environment, then check how the destination domain is resolved. Using an IP for the server doesn't eliminate the need to resolve destination domains. For a connection timeout, consider the network route to the server, port reachability, and service status. A timeout isn't specific to an incorrect password; don't delete all your configuration based on one error.

If the error clearly indicates authentication, check the original details for the password, UUID, and other fields. Make sure you didn't enter the display name as a username or copy extra spaces. For TLS errors, check the device clock, Server Name, and certificate requirements. Narrow the issue down by error category instead of randomly changing encryption settings. If details are missing, ask your existing service provider.

If a webpage loads but some images or sign-in steps fail, treat the page as multiple requests rather than a single destination. The main page, static resources, authentication, and APIs may use different domains and match different rules. Check available connection logs for the failing destination and its policy, then decide whether to add a rule. Don't route all traffic through the same path indefinitely just because one image failed.

Use controlled comparisons to identify the cause

If you suspect a routing rule, keep the original configuration and temporarily compare Proxy with Config using the same destination and a new connection. If you suspect the network route, keep the server and configuration unchanged and switch to another network you can use. Change one condition at a time, record the result, then restore it. This helps identify whether the problem is in the configuration, service, or network.

Some apps keep long-lived connections and may continue using an old session after you change a rule. To test a new rule, close the relevant page or restart the app. If needed, reconnect Shadowrocket before sending a new request. Browser cache may also show old content, so a page that appears to load doesn't necessarily prove that a new network request succeeded. Test with content that clearly needs to be fetched online.

If the cause is still unclear, prepare a concise problem report: device and network type, steps to reproduce, type of destination, Global Routing mode, exact error, and single-variable tests already tried. Remove full subscriptions, passwords, QR codes, and credentials before sharing. Check for the same symptom in the FAQ first. Repeated imports can make the original problem harder to diagnose.

06 / Data & Traffic

Understand traffic stats and request logs

Data is not a billing page

Data shows the traffic statistics and related records available in the app. Before interpreting a number, check whether it applies to the current connection, a particular period, or a cumulative total, and note the units and upload/download direction. Different screens may use different ranges; similar numbers don't necessarily mean the stats are identical. If the interface offers a reset or clear option, check whether it affects displayed history or other data before using it.

Traffic shown by the client, system cellular usage, and usage records from your existing service may be measured in different places and under different rules. Protocol overhead, retransmissions, background connections, direct traffic, and billing periods can all cause differences. A change in Data only shows that traffic occurred within the relevant measurement scope. It doesn't prove that the remote service successfully handled every request, and it isn't a balance for your service plan.

To compare traffic from a specific action, note the starting value and time period, stop unnecessary background tasks, perform one clearly defined action, then record the ending value. Don't play media and download files at the same time and then attribute the change to a webpage you just opened. Before resetting a counter, save the current record. Resetting the app's display doesn't change system statistics or your service billing period.

Use request destinations to check routing

If the app provides connection or request logs, focus on the destination domain or address, matched rule, selected policy, and error. These can help answer whether a request passed through the client, which condition it matched, and which route it used. Not every log entry includes all of this information. If a domain is missing, consider the request type and resolution process rather than guessing the domain.

A webpage can generate many connections, including fonts, images, video segments, and background APIs. The number of log entries isn't the number of times you opened a page. An app may also keep syncing after the screen turns off. To troubleshoot, narrow the time range to the action you just performed, then review destinations and policies. Don't assume every unfamiliar domain in the list is a fault or suspicious activity.

In Config mode, if a destination matched DIRECT, check rule order and address conditions first instead of changing proxy servers again. If it matched REJECT, confirm that blocking the connection was intended. If the log points to a custom policy group, check which target is actually selected in that group. Seeing the group name doesn't mean every server in it was used.

ObservationWhat it tells youWhat to check next
Counter increasesData was transferred within the measured scope.Whether the task completed and the route was expected.
Destination appears in logsThe app observed a related connection.Matched rule, actual policy, and cause of any error.
Log shows REJECTA rule actively blocked the connection.Whether the rule is blocking a feature by mistake.

Use logs for short-term troubleshooting

If a destination you just used doesn't appear in the logs, first check whether logging is enabled and whether filters are hiding entries. Then check if the destination reused an old connection or didn't make a network request at all. A browser may load a page from cache without generating the expected entry. No log entry isn't proof that a rule failed; start with a clear, new request for comparison.

For a specific issue, temporarily enable the detailed logging you need, reproduce it once, save the relevant excerpt, then restore the previous logging level. Keeping extensive diagnostic data for a long time makes it harder to find information and may record personal browsing destinations. Before exporting, include only the relevant time period and remove credentials, private domains, and personally identifying parameters. You don't need to share your entire usage history.

Before clearing Data records, distinguish between clearing history, resetting counters, and deleting connection details. These actions have different effects. If the interface wording is unclear, cancel and check before proceeding. Clearing all history to fix a webpage usually won't correct server settings and may erase useful troubleshooting evidence. Keep the evidence first, then make a targeted change.

Start with a specific question, then choose the relevant record: check traffic stats to see whether data was transferred; rules and policies to see which route a destination took; error messages to identify the failure layer; and your existing service records to check usage balances. Keeping these measures separate makes Data useful for verifying configuration instead of treating it as a score for connection quality.

07 / Common Settings

Adjust Settings for your needs

When to use On Demand versus manual connections

Adjust Settings for your actual needs; you don't have to enable every option. For initial setup, keep only the settings needed for verification. Add automatic connection or diagnostic options after servers, rules, and destination access are working. Menu locations may vary by system and app version, so look for the function rather than assuming the setting mentioned here appears under the same name on your screen.

On Demand triggers connection behavior under conditions supported by the system; it doesn't make a connection faster. Before enabling it, make sure manual connections work reliably, then review the available network conditions and exceptions one at a time. If manual connections fail because of incorrect settings, automatic triggers will just repeat the failure. Don't use On Demand to hide an unresolved disconnection, or assume that automatic reconnection means a server is always available.

When setting up on-demand behavior, check what happens when switching between Wi-Fi and cellular, and whether you need the connection off on certain networks. After saving the conditions, test once on a network that meets them and once on one that doesn't. If the connection turns back on after you switch it off manually, check whether an On Demand condition is still triggering before deleting the VPN configuration or reinstalling the app.

DNS, connection tests, and configuration sources

DNS settings may be in Settings or in the active Config. First identify which one is in effect. DNS resolves domain names to addresses; proxy policies determine the connection route. They are related but different. After changing DNS settings, cached results and old connections may persist, so test with a new request. Don't change DNS and replace the entire rules configuration at the same time, or it will be difficult to tell what helped.

For a domain resolution issue, check the spelling, current network, and whether the configured DNS server is reachable. Then check whether a policy affects DNS requests. Cycling through every DNS option isn't a reliable way to reach a conclusion. If the configuration references remote rules or other resources, check that they can be fetched and what happens when fetching fails. Don't overwrite a working file without understanding the result.

Connectivity Test results also depend on the test method and target. After changing the target, don't directly compare the new result with a number measured under different conditions. To check whether a server can reach a real service, use consistent conditions for a probe, then test the actual service. Record the protocol, server region, and network time so changes in the numbers have context.

Handle certificates, backups, and advanced options carefully

Ordinary server connections and rule-based routing don't require extra certificates just to make the setup feel complete. Features that decrypt HTTPS change how the device trusts affected connections. Evaluate these separately after considering their purpose, scope, and privacy impact; they're not required for a first-time setup. If you're unsure what a certificate option does, leave it unchanged rather than approving every prompt from an unfamiliar configuration.

If sync or backup is enabled, check which account is used, what gets synced, and where sensitive information is stored. Configuration files may contain server credentials, and subscription URLs may grant access, so treat backups as private. A screenshot can record the layout but isn't a suitable backup for full credentials. Don't rely on public cloud-sharing links for routine personal configuration backups.

Adjust UDP, IPv6, routing exclusions, and other connection options based on your existing service's capabilities and the actual issue. One app needing a particular transport doesn't mean every server supports it. A network offering IPv6 doesn't guarantee that every address rule in the configuration covers it. Check the feature requirements before changing a switch so you don't mistake a compatibility issue for an installation problem.

After changing settings, keep a simple change log with the setting name, reason, test destination, and rollback method. Don't include full credentials. This helps you distinguish defaults from your own exceptions after an app update or device migration. For interface terminology, see the Quick Reference. For a specific issue, return to the connection verification chapter.

08 / Ongoing Maintenance

Updates, backups, and device migration

Maintain the app, subscriptions, and Config separately

Treat maintenance as three separate tasks: app updates, subscription updates, and Config updates. App updates change client features and compatibility; subscription updates change your existing server list; Config updates may change rules, policies, and DNS settings. You don't need to do all three at once. When connections are working, preserve that state and make changes only when you know what's being updated and what it affects.

Manage iPhone and iPad app updates through the App Store. For Mac, Windows, Android, and Linux, download an installer for your system from this site's download page. The iPhone and iPad app costs $2.99 as a one-time App Store purchase; versions for Mac, Windows, Android, and Linux are free. Back up your details and stop important active connections before updating. Don't treat the time a file finishes downloading as the product's release date.

Before updating a subscription, note the current entries and groups. Afterward, check whether the selected destination is still there. Server details may change even when an entry keeps the same name, so don't assume nothing changed based on the name alone. If an automatic update fails, check the URL and current access conditions before deleting the entire group. Deleting and reimporting may lose your selections and troubleshooting notes.

Make sure backups can be restored

Before backing up, list what you need to keep: manually added servers, subscription URLs, the active Config, your rule changes, policy group selections, and common settings. Export methods may cover different things; a rules file may not include subscription details, and App Store purchase history doesn't save personal connection settings. After exporting, make sure the file opens and contains what you expect, then store it somewhere private that you control.

If your current version and device support iCloud sync, check the relevant system account and sync permissions, then confirm which items the app actually syncs. Don't assume migration is ready just because a switch is on. Before an important migration, keep a separate export and note the name of the active configuration on the old device. Sync can propagate accidental deletions and bad changes; a separate copy lets you return to the pre-migration state.

Treat backup files and subscription URLs as credentials. Don't put passwords in filenames, full URLs in public notes, or configurations with identity fields in public discussions. If you think a subscription URL has been exposed, contact your existing service provider to replace the associated access credentials, then update the old details in the app and your backups. Deleting a chat message can't take back content someone already saved.

Restore in dependency order when changing devices

On a new iPhone or iPad, first restore the app from Purchased under the original account, then open it and grant permissions as needed. Next restore servers and subscriptions, then import Config. Finally, set Global Routing, policy selections, and On Demand. Don't enable policies that reference servers before restoring them, or restore every automatic trigger before basic connection tests pass.

After restoring, repeat the basic network, server, and destination checks in chapter five. Then review LAN exceptions and everyday rules. A new device may have different system permissions, and a new network may use different address ranges; successful tests on the old device don't replace testing on the new one. Confirm everything works and keep a backup before deciding whether to clear the old device, so you don't lose information before restoration is complete.

For cross-platform migration, handle service details separately from platform-specific configuration features. First check which import formats and settings the destination platform supports, then restore items individually. Don't assume every system option exported from a mobile device will behave the same way on a computer. Mac users should also check the installer matches their chip; Linux users should keep to the package format for their distribution instead of installing several formats over one another.

If a problem appears after an update, compare the individual changes and try restoring the backed-up configuration or settings before clearing the app. When asking for help, provide reproducible steps, exact error text, and comparisons you've already run; remove personal credentials. For export and restoration steps, see Configuration Backup and Device Migration. To find recurring connection issues by category, see the FAQ.

Maintenance is complete when you can explain and restore the current setup, not just when there are no warnings. Know where you got the app, which existing service each subscription belongs to, which configuration is active, how common destinations are routed, and where backups are stored. Keeping these details easy to check makes it possible to repeat the same verification steps after changing networks, updating the app, or moving to another device.

Download Shadowrocket for all platforms