oimwiodev/chrome-use
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

add iOS support (#358)

Browse Source 
* ios

* tests

* docs

* real device

* better list

* fixes
This commit is contained in:
Chris Tate
2026-02-03 01:36:19 -06:00
committed by GitHub
parent 3c6ae7df9d
commit a1b992411e
18 changed files with 4197 additions and 56 deletions
Download Patch File Download Diff File Expand all files Collapse all files
cli/src/commands.rs
+49
View File
@@ -103,6 +103,12 @@ pub fn parse_command(args: &[String], flags: &Flags) -> Result<Value, ParseError
nav_cmd["headers"] = headers;
}
}
// Include iOS device info if specified (needed for auto-launch with existing daemon)
if flags.provider.as_deref() == Some("ios") {
if let Some(ref device) = flags.device {
nav_cmd["iosDevice"] = json!(device);
}
}
Ok(nav_cmd)
}
"back" => Ok(json!({ "id": id, "action": "back" })),
@@ -835,6 +841,48 @@ pub fn parse_command(args: &[String], flags: &Flags) -> Result<Value, ParseError
}
}
// === iOS-specific commands ===
"tap" => {
// Alias for click (semantic clarity for touch interfaces)
let sel = rest.get(0).ok_or_else(|| ParseError::MissingArguments {
context: "tap".to_string(),
usage: "tap <selector>",
})?;
Ok(json!({ "id": id, "action": "tap", "selector": sel }))
}
"swipe" => {
let direction = rest.get(0).ok_or_else(|| ParseError::MissingArguments {
context: "swipe".to_string(),
usage: "swipe <up|down|left|right> [distance]",
})?;
let valid_directions = ["up", "down", "left", "right"];
if !valid_directions.contains(direction) {
return Err(ParseError::InvalidValue {
message: format!("Invalid swipe direction: {}", direction),
usage: "swipe <up|down|left|right> [distance]",
});
}
let mut cmd = json!({ "id": id, "action": "swipe", "direction": direction });
if let Some(distance) = rest.get(1) {
if let Ok(d) = distance.parse::<u32>() {
cmd.as_object_mut().unwrap().insert("distance".to_string(), json!(d));
}
}
Ok(cmd)
}
"device" => {
match rest.get(0).map(|s| *s) {
Some("list") | None => {
// List available iOS simulators
Ok(json!({ "id": id, "action": "device_list" }))
}
Some(sub) => Err(ParseError::UnknownSubcommand {
subcommand: sub.to_string(),
valid_options: &["list"],
}),
}
}
_ => Err(ParseError::UnknownCommand {
command: cmd.to_string(),
}),
@@ -1376,6 +1424,7 @@ mod tests {
user_agent: None,
provider: None,
ignore_https_errors: false,
device: None,
}
}
cli/src/connection.rs
+18
View File
@@ -215,6 +215,8 @@ pub fn ensure_daemon(
ignore_https_errors: bool,
profile: Option<&str>,
state: Option<&str>,
provider: Option<&str>,
device: Option<&str>,
) -> Result<DaemonResult, String> {
// Check if daemon is running AND responsive
if is_daemon_running(session) && daemon_ready(session) {
@@ -343,6 +345,14 @@ pub fn ensure_daemon(
cmd.env("AGENT_BROWSER_STATE", st);
}
if let Some(p) = provider {
cmd.env("AGENT_BROWSER_PROVIDER", p);
}
if let Some(d) = device {
cmd.env("AGENT_BROWSER_IOS_DEVICE", d);
}
// Create new process group and session to fully detach
unsafe {
cmd.pre_exec(|| {
@@ -410,6 +420,14 @@ pub fn ensure_daemon(
cmd.env("AGENT_BROWSER_STATE", st);
}
if let Some(p) = provider {
cmd.env("AGENT_BROWSER_PROVIDER", p);
}
if let Some(d) = device {
cmd.env("AGENT_BROWSER_IOS_DEVICE", d);
}
// CREATE_NEW_PROCESS_GROUP | DETACHED_PROCESS
const CREATE_NEW_PROCESS_GROUP: u32 = 0x00000200;
const DETACHED_PROCESS: u32 = 0x00000008;
cli/src/flags.rs
+9
View File
@@ -18,6 +18,7 @@ pub struct Flags {
pub user_agent: Option<String>,
pub provider: Option<String>,
pub ignore_https_errors: bool,
pub device: Option<String>,
}
pub fn parse_flags(args: &[String]) -> Flags {
@@ -49,6 +50,7 @@ pub fn parse_flags(args: &[String]) -> Flags {
user_agent: env::var("AGENT_BROWSER_USER_AGENT").ok(),
provider: env::var("AGENT_BROWSER_PROVIDER").ok(),
ignore_https_errors: false,
device: env::var("AGENT_BROWSER_IOS_DEVICE").ok(),
};
let mut i = 0;
@@ -131,6 +133,12 @@ pub fn parse_flags(args: &[String]) -> Flags {
}
}
"--ignore-https-errors" => flags.ignore_https_errors = true,
"--device" => {
if let Some(d) = args.get(i + 1) {
flags.device = Some(d.clone());
i += 1;
}
}
_ => {}
}
i += 1;
@@ -165,6 +173,7 @@ pub fn clean_args(args: &[String]) -> Vec<String> {
"--user-agent",
"-p",
"--provider",
"--device",
];
for arg in args.iter() {
cli/src/main.rs
+2
View File
@@ -207,6 +207,8 @@ fn main() {
flags.ignore_https_errors,
flags.profile.as_deref(),
flags.state.as_deref(),
flags.provider.as_deref(),
flags.device.as_deref(),
) {
Ok(result) => result,
Err(e) => {
cli/src/output.rs
+121 -2
View File
@@ -78,6 +78,53 @@ pub fn print_response(resp: &Response, json_mode: bool, action: Option<&str>) {
);
return;
}
// iOS Devices
if let Some(devices) = data.get("devices").and_then(|v| v.as_array()) {
if devices.is_empty() {
println!("No iOS devices available. Open Xcode to download simulator runtimes.");
return;
}
// Separate real devices from simulators
let real_devices: Vec<_> = devices
.iter()
.filter(|d| d.get("isRealDevice").and_then(|v| v.as_bool()).unwrap_or(false))
.collect();
let simulators: Vec<_> = devices
.iter()
.filter(|d| !d.get("isRealDevice").and_then(|v| v.as_bool()).unwrap_or(false))
.collect();
if !real_devices.is_empty() {
println!("Connected Devices:\n");
for device in real_devices.iter() {
let name = device.get("name").and_then(|v| v.as_str()).unwrap_or("Unknown");
let runtime = device.get("runtime").and_then(|v| v.as_str()).unwrap_or("");
let udid = device.get("udid").and_then(|v| v.as_str()).unwrap_or("");
println!(" {} {} ({})", color::green("●"), name, runtime);
println!(" {}", color::dim(udid));
}
println!();
}
if !simulators.is_empty() {
println!("Simulators:\n");
for device in simulators.iter() {
let name = device.get("name").and_then(|v| v.as_str()).unwrap_or("Unknown");
let runtime = device.get("runtime").and_then(|v| v.as_str()).unwrap_or("");
let state = device.get("state").and_then(|v| v.as_str()).unwrap_or("Unknown");
let udid = device.get("udid").and_then(|v| v.as_str()).unwrap_or("");
let state_indicator = if state == "Booted" {
color::green("●")
} else {
color::dim("○")
};
println!(" {} {} ({})", state_indicator, name, runtime);
println!(" {}", color::dim(udid));
}
}
return;
}
// Tabs
if let Some(tabs) = data.get("tabs").and_then(|v| v.as_array()) {
for (i, tab) in tabs.iter().enumerate() {
@@ -1551,6 +1598,68 @@ Examples:
"##
}
// === iOS Commands ===
"tap" => {
r##"
agent-browser tap - Tap an element (touch gesture)
Usage: agent-browser tap <selector>
Taps an element. This is an alias for 'click' that provides semantic clarity
for touch-based interfaces like iOS Safari.
Options:
--json Output as JSON
--session <name> Use specific session
Examples:
agent-browser tap "#submit-button"
agent-browser tap @e1
agent-browser -p ios tap "button:has-text('Sign In')"
"##
}
"swipe" => {
r##"
agent-browser swipe - Swipe gesture (iOS)
Usage: agent-browser swipe <direction> [distance]
Performs a swipe gesture on iOS Safari. The direction determines
which way the content moves (swipe up scrolls down, etc.).
Arguments:
direction up, down, left, or right
distance Optional distance in pixels (default: 300)
Options:
--json Output as JSON
--session <name> Use specific session
Examples:
agent-browser -p ios swipe up
agent-browser -p ios swipe down 500
agent-browser -p ios swipe left
"##
}
"device" => {
r##"
agent-browser device - Manage iOS simulators
Usage: agent-browser device <subcommand>
Subcommands:
list List available iOS simulators
Options:
--json Output as JSON
--session <name> Use specific session
Examples:
agent-browser device list
agent-browser -p ios device list
"##
}
_ => return false,
};
println!("{}", help.trim());
@@ -1660,7 +1769,8 @@ Options:
--proxy-bypass <hosts> Bypass proxy for these hosts (or AGENT_BROWSER_PROXY_BYPASS)
e.g., --proxy-bypass "localhost,*.internal.com"
--ignore-https-errors Ignore HTTPS certificate errors
-p, --provider <name> Cloud browser provider (or AGENT_BROWSER_PROVIDER env)
-p, --provider <name> Browser provider: ios, browserbase, kernel, browseruse
--device <name> iOS device name (e.g., "iPhone 15 Pro")
--json JSON output
--full, -f Full page screenshot
--headed Show browser window (not headless)
@@ -1671,8 +1781,10 @@ Options:
Environment:
AGENT_BROWSER_SESSION Session name (default: "default")
AGENT_BROWSER_EXECUTABLE_PATH Custom browser executable path
AGENT_BROWSER_PROVIDER Cloud browser provider
AGENT_BROWSER_PROVIDER Browser provider (ios, browserbase, kernel, browseruse)
AGENT_BROWSER_STREAM_PORT Enable WebSocket streaming on port (e.g., 9223)
AGENT_BROWSER_IOS_DEVICE Default iOS device name
AGENT_BROWSER_IOS_UDID Default iOS device UDID
Examples:
agent-browser open example.com
@@ -1684,6 +1796,13 @@ Examples:
agent-browser screenshot --full
agent-browser --cdp 9222 snapshot # Connect via CDP port
agent-browser --profile ~/.myapp open example.com # Persistent profile
iOS Simulator (requires Xcode and Appium):
agent-browser -p ios open example.com # Use default iPhone
agent-browser -p ios --device "iPhone 15 Pro" open url # Specific device
agent-browser -p ios device list # List simulators
agent-browser -p ios swipe up # Swipe gesture
agent-browser -p ios tap @e1 # Touch element
"#
);
}