Official Guide · Glossary

Shadowrocket
App, protocol, and rule terms

Identify which layer a term belongs to before changing settings. App screens determine where to make changes, protocols determine how connections work, and rules determine the path a request takes. A single test only reflects part of the connection at that moment.

This page keeps app interface terms, protocol names, and rule keywords in English. Similar names do not mean the same thing: for example, Proxy in Global Routing is an overall routing mode, while PROXY at the end of a rule must be understood in the context of the current configuration. After browsing the categories, see Getting Started for step-by-step instructions.

App & Connection Settings

Screens, routing modes, and connection behavior

First identify whether you are changing the overall mode, a server entry, or the conditions for automatic connections. Screen locations can vary by platform; use the current app interface as your reference.

Home

Home is where you view the server list, select a server, and toggle the connection. Start by confirming which entry is selected. A server in the list or a selection indicator does not mean a connection has been established. Also check the connection toggle and system connection status, then verify by accessing your intended destination.

Global Routing

Global Routing determines the overall routing mode. The options are Config, Proxy, and Direct. Config handles requests according to the current configuration, Proxy sends applicable requests through the selected proxy policy, and Direct connects without a proxy. If rules are not working, first check that this is set to Config, then review the specific rules.

On Demand

On Demand requests a connection when preset conditions are met; it is not a server performance setting. Whether it triggers depends on system permissions, network conditions, and the configured conditions. The name alone does not mean it will reconnect automatically in every situation. If the connection returns after you disconnect manually, check these settings first.

Add Server

Add Server lets you enter server details manually. Typically, you select a type first, then enter the address, port, and authentication fields. The protocol type must match the remote server; do not switch types at random to test speed. After entering the details, check the transport method, TLS, server name, and other parameters as well.

Settings

Settings groups options that control client behavior; it is separate from the edit screen for an individual server. Before troubleshooting, note the original values and change one option at a time, then test the same destination again. Options may differ between versions and platforms, so do not assume another interface uses the same layout.

Protocols

Server types and transport requirements

Choose a protocol type based on the details provided by your service provider. A protocol name alone says nothing about connection quality and cannot replace the complete connection parameters.

Shadowsocks

Shadowsocks is a proxy protocol that uses pre-shared credentials. The server address, port, encryption method, and password must all match. Even if the address resolves, a mismatched password or encryption method can prevent communication. Keep the original parameters after importing; do not mix fields from different server entries.

VMess

VMess uses a user ID and other parameters for authentication and may use different transport methods. Selecting the right server type only confirms the protocol layer; the transport path, Host, TLS, and other fields must also match. If an imported server will not connect, compare every field with the original details rather than checking only the user ID.

VLESS

VLESS is a lightweight proxy protocol that authenticates with a user ID but does not encrypt content itself, so consider it alongside transport security settings. Its type and required fields differ from VMess; changing the type alone does not convert a configuration. Set TLS and the transport method to match the remote configuration.

Trojan

Trojan connections typically use TLS. The password is used for authentication, while the server name and certificate verification are separate items to check. A reachable address does not guarantee a successful TLS handshake; a name mismatch or incorrect device time may also cause problems. If you see a certificate error, check its cause before changing verification settings.

Hysteria2

Hysteria2 is a QUIC-based proxy protocol, so check whether UDP is reachable on the current network. Correct authentication and TLS parameters do not guarantee the same connection will work on every network. If results differ after switching networks, test the same entry on both and check whether network conditions are responsible.

WireGuard

WireGuard is a key-pair-based tunneling protocol. Its configuration typically includes an interface address, peer public key, endpoint, and allowed address ranges. The endpoint locates the peer, but the keys and routing ranges also affect connectivity. A private key is sensitive information; redact it from troubleshooting materials and never share it as a public example.

HTTP

As a server type, HTTP means an HTTP proxy; it does not mean entering a webpage URL in the server field. Supported requests and whether a username and password are required depend on the server and its configuration. Accessing an HTTPS destination does not automatically mean traffic between the client and proxy server is encrypted.

SOCKS5

SOCKS5 is a proxy protocol that forwards connection requests. Check authentication requirements and UDP support on both the client and server. It is not itself an encrypted tunnel, so the protocol name alone does not indicate transport security. When troubleshooting, check authentication, destination connectivity, and UDP support separately.

Rule Keywords

Match conditions and handling policies

Rules are generally checked from top to bottom, and the matching rule's policy is applied. The snippets below explain the syntax; they are not a complete configuration to replace your existing settings.

DOMAIN

DOMAIN matches an exact domain name, making it suitable for a single, specific hostname. Enter the domain itself without a protocol prefix, webpage path, or query parameters. If the destination uses a different subdomain, this exact-match rule may not apply. Check the connection log to confirm the requested hostname.

DOMAIN,example.com,DIRECT

DOMAIN-SUFFIX

DOMAIN-SUFFIX matches a domain suffix and is commonly used to cover a domain and its subdomains. Its scope is broader than an exact domain match, so confirm that related subdomains should use the same policy before adding the rule. If a rule appears correct but does not take effect, check whether an earlier rule matched first.

DOMAIN-SUFFIX,example.com,PROXY

DOMAIN-KEYWORD

DOMAIN-KEYWORD matches a keyword in the domain name, not text on the webpage. Short or common keywords may match unrelated domains, so check the scope before using them. When DOMAIN or DOMAIN-SUFFIX can express the match precisely, those rules are usually easier to review and maintain.

DOMAIN-KEYWORD,example,DIRECT

GEOIP

GEOIP matches according to the region associated with an IP address in a geolocation database; it does not read the visitor's actual location. Database records, resolved destination addresses, and changes to service infrastructure can all affect the result. If an unexpected rule matches, check the actual destination IP instead of relying on the website name.

GEOIP,CN,DIRECT

IP-CIDR

IP-CIDR matches IPv4 address ranges in CIDR notation. The number after the slash is the network prefix length: a longer prefix usually covers a smaller range. It is not a port number or priority. When maintaining these rules, check both the address range and whether the destination actually uses IPv4.

IP-CIDR,192.0.2.0/24,DIRECT

IP-CIDR6

IP-CIDR6 matches IPv6 address ranges. Its purpose is similar to IPv4 range rules, but the address format is different. If a destination supports both IPv4 and IPv6, changing IP-CIDR alone may not cover every connection. To troubleshoot differences on dual-stack networks, first check which address family the request uses.

IP-CIDR6,2001:db8::/32,DIRECT

FINAL

FINAL is the fallback rule used when none of the preceding rules match, and is usually placed at the end of the rule list. If it comes before specific matches, later rules may never be checked. Changing its policy affects all requests not covered by earlier rules, so decide on the intended default route first.

FINAL,DIRECT

PROXY / DIRECT / REJECT

These three names are commonly used at the end of rules: PROXY usually refers to a proxy policy, DIRECT means a direct connection, and REJECT means the request is denied. Check the policy definition in the current configuration to see what PROXY actually points to; the name alone does not identify a specific server. Do not confuse uppercase rule names with the Proxy and Direct options in Global Routing.

URL Rewrite

URL Rewrite rewrites request URLs according to matching conditions. It is separate from selecting a proxy server or changing overall routing. Whether a rule matches depends on its syntax and the request content available; HTTPS scenarios may involve additional requirements. Do not enable rewrite rules you do not understand to troubleshoot an ordinary connection problem.

Subscriptions & Configuration

Server sources and configuration files

The server list answers “where to connect”; a configuration file answers “how to handle requests.” Updating one does not necessarily change the other.

Subscribe

Subscribe retrieves and updates server entries from a subscription URL. Get the link from your service provider. An update may add, remove, or change entries, and can overwrite manual edits. Subscription URLs often contain access credentials, so redact the full address and authentication parameters before sharing screenshots or logs.

Server

A server entry stores the details needed for a remote connection and is also commonly called a node. It usually includes an address, port, protocol, and authentication fields. The display name is just a label; it does not prove the server's location, speed, or connection status. To tell whether two entries are the same, compare their actual parameters, not just their names.

Config

Config can mean a configuration file containing rules, policies, and related settings, or the Global Routing option that handles requests according to the configuration. After importing a file, check which configuration is currently active; having the file does not mean it is in use. Keep the original file before switching so you can restore it or compare rule changes.

iCloud Sync

iCloud Sync uses iCloud to sync app-supported data across supported devices. Check the in-app settings for the exact scope. It is not the same as restoring purchases and does not mean all system permissions transfer automatically. After moving to a new device, check the servers, active configuration, and connection permissions separately; do not rely on the server list alone.

Connections & Diagnostics

Resolution, tests, and real-world access

Separate the stages where a failure can occur. Successful DNS resolution, a passing test, and access to a destination are related, but they are not equivalent.

DNS

DNS resolves domain names to IP addresses, a step required before many connections can be made. Successful resolution does not guarantee the connection will succeed, and a resolution failure does not necessarily indicate a server authentication problem. Record the DNS settings, resolved address, and subsequent error separately rather than changing every related option at once.

Latency

Latency is the time taken by a particular test. Its meaning depends on the test method, and it cannot be directly converted to sustained download speed. To compare servers, use the same network and test method, and repeat the tests at roughly the same time. A single timeout only means that test did not finish; check actual access as well.

Connectivity Test

Connectivity Test is the test entry point for checking server connectivity. Interpret results in light of the selected test method. A successful test does not guarantee that every website or request type follows the same path, nor does it prove that the rules are configured correctly. If the test passes but access fails, continue by checking DNS, rule matches, and destination connection errors.

Connection Log

The connection log records request times, destinations, rule matches, and errors. Check it right after reproducing a problem, focusing on entries from the same time period. Distinguish between DNS resolution, authentication, TLS, and destination timeouts. Before sharing logs, redact passwords, user IDs, subscription credentials, and personal browsing data.

Record troubleshooting results in the same order

First note the current network and Global Routing mode, then confirm the selected server, active Config, and actual request destination. Change one condition at a time and record the result after retesting. This helps distinguish the effect of a parameter change from a coincidental network change. To learn how VMess and VLESS fields differ, read VMess vs. VLESS: Key Differences.