fix: harden Lightpanda startup timeouts (#762)

cli/src/native/browser.rs

@@ -1,12 +1,13 @@
 use serde_json::{json, Value};
 use std::collections::HashSet;
+use std::future::Future;
 use std::sync::Arc;
+use std::time::{Duration, Instant};
 use tokio::sync::{broadcast, Mutex};
 
-use super::cdp::chrome::{
-    auto_connect_cdp, discover_cdp_url, launch_chrome, ChromeProcess, LaunchOptions,
-};
+use super::cdp::chrome::{auto_connect_cdp, launch_chrome, ChromeProcess, LaunchOptions};
 use super::cdp::client::CdpClient;
+use super::cdp::discovery::discover_cdp_url;
 use super::cdp::lightpanda::{launch_lightpanda, LightpandaLaunchOptions, LightpandaProcess};
 use super::cdp::types::*;
 
@@ -167,6 +168,10 @@ pub struct BrowserManager {
     default_timeout_ms: u64,
 }
 
+const LIGHTPANDA_CDP_CONNECT_TIMEOUT: Duration = Duration::from_secs(5);
+const LIGHTPANDA_CDP_CONNECT_POLL_INTERVAL: Duration = Duration::from_millis(100);
+const LIGHTPANDA_TARGET_INIT_TIMEOUT: Duration = Duration::from_secs(10);
+
 impl BrowserManager {
     pub async fn launch(options: LaunchOptions, engine: Option<&str>) -> Result<Self, String> {
         let engine = engine.unwrap_or("chrome");
@@ -205,9 +210,7 @@ impl BrowserManager {
                     proxy: options.proxy.clone(),
                     port: None,
                 };
-                let lp = tokio::task::spawn_blocking(move || launch_lightpanda(&lp_options))
-                    .await
-                    .map_err(|e| format!("Lightpanda launch task failed: {}", e))??;
+                let lp = launch_lightpanda(&lp_options).await?;
                 let url = lp.ws_url.clone();
                 (url, BrowserProcess::Lightpanda(lp))
             }
@@ -220,18 +223,22 @@ impl BrowserManager {
             }
         };
 
-        let client = CdpClient::connect(&ws_url).await?;
-        let mut manager = Self {
-            client,
-            browser_process: Some(process),
-            ws_url,
-            pages: Vec::new(),
-            active_page_index: 0,
-            default_timeout_ms: 25_000,
+        let manager = if engine == "lightpanda" {
+            initialize_lightpanda_manager(ws_url, process).await?
+        } else {
+            let client = CdpClient::connect(&ws_url).await?;
+            let mut manager = Self {
+                client,
+                browser_process: Some(process),
+                ws_url,
+                pages: Vec::new(),
+                active_page_index: 0,
+                default_timeout_ms: 25_000,
+            };
+            manager.discover_and_attach_targets().await?;
+            manager
         };
 
-        manager.discover_and_attach_targets().await?;
-
         let session_id = manager.active_session_id()?.to_string();
 
         if ignore_https_errors {
@@ -1119,6 +1126,120 @@ impl BrowserManager {
     }
 }
 
+async fn connect_cdp_with_retry(
+    ws_url: &str,
+    total_timeout: Duration,
+    poll_interval: Duration,
+) -> Result<CdpClient, String> {
+    let deadline = Instant::now() + total_timeout;
+
+    loop {
+        match CdpClient::connect(ws_url).await {
+            Ok(client) => return Ok(client),
+            Err(err) => {
+                if Instant::now() >= deadline {
+                    return Err(err);
+                }
+            }
+        }
+
+        tokio::time::sleep(poll_interval).await;
+    }
+}
+
+async fn initialize_lightpanda_manager(
+    ws_url: String,
+    process: BrowserProcess,
+) -> Result<BrowserManager, String> {
+    let deadline = Instant::now() + LIGHTPANDA_TARGET_INIT_TIMEOUT;
+    let mut process = Some(process);
+
+    loop {
+        let client = match connect_cdp_with_retry(
+            &ws_url,
+            LIGHTPANDA_CDP_CONNECT_TIMEOUT,
+            LIGHTPANDA_CDP_CONNECT_POLL_INTERVAL,
+        )
+        .await
+        {
+            Ok(client) => client,
+            Err(err) => {
+                if Instant::now() >= deadline {
+                    return Err(lightpanda_target_init_timeout(Some(&err)));
+                }
+                tokio::time::sleep(LIGHTPANDA_CDP_CONNECT_POLL_INTERVAL).await;
+                continue;
+            }
+        };
+
+        let mut manager = BrowserManager {
+            client,
+            browser_process: None,
+            ws_url: ws_url.clone(),
+            pages: Vec::new(),
+            active_page_index: 0,
+            default_timeout_ms: 25_000,
+        };
+
+        match discover_and_attach_lightpanda_targets(&mut manager, deadline).await {
+            Ok(()) => {
+                manager.browser_process = process.take();
+                return Ok(manager);
+            }
+            Err(err) => {
+                if Instant::now() >= deadline {
+                    return Err(lightpanda_target_init_timeout(Some(&err)));
+                }
+                tokio::time::sleep(LIGHTPANDA_CDP_CONNECT_POLL_INTERVAL).await;
+            }
+        }
+    }
+}
+
+async fn discover_and_attach_lightpanda_targets(
+    manager: &mut BrowserManager,
+    deadline: Instant,
+) -> Result<(), String> {
+    run_with_lightpanda_deadline(
+        deadline,
+        manager.discover_and_attach_targets(),
+        "Target domain initialization attempt exceeded the remaining startup deadline",
+    )
+    .await
+}
+
+fn remaining_until(deadline: Instant) -> Option<Duration> {
+    deadline.checked_duration_since(Instant::now())
+}
+
+async fn run_with_lightpanda_deadline<F, T>(
+    deadline: Instant,
+    operation: F,
+    timeout_context: &'static str,
+) -> Result<T, String>
+where
+    F: Future<Output = Result<T, String>>,
+{
+    let remaining = remaining_until(deadline)
+        .ok_or_else(|| lightpanda_target_init_timeout(Some("deadline expired before retry")))?;
+
+    match tokio::time::timeout(remaining, operation).await {
+        Ok(result) => result,
+        Err(_) => Err(lightpanda_target_init_timeout(Some(timeout_context))),
+    }
+}
+
+fn lightpanda_target_init_timeout(last_error: Option<&str>) -> String {
+    let mut message = format!(
+        "Timed out after {}ms waiting for Lightpanda Target domain to initialize",
+        LIGHTPANDA_TARGET_INIT_TIMEOUT.as_millis(),
+    );
+    if let Some(last_error) = last_error {
+        message.push_str(&format!("\nLast error: {}", last_error));
+    }
+    message
+}
+
 async fn resolve_cdp_url(input: &str) -> Result<String, String> {
     if input.starts_with("ws://") || input.starts_with("wss://") {
         return Ok(input.to_string());
@@ -1145,6 +1266,7 @@ async fn resolve_cdp_url(input: &str) -> Result<String, String> {
 #[cfg(test)]
 mod tests {
     use super::*;
+    use tokio::time::sleep;
 
     #[test]
     fn test_validate_launch_options_extensions_and_cdp() {
@@ -1251,4 +1373,59 @@ mod tests {
             "Element not found. Verify the selector is correct and the element exists in the DOM.";
         assert_eq!(to_ai_friendly_error("No element found for css 'x'"), mapped);
     }
+
+    #[test]
+    fn test_remaining_until_returns_none_for_past_deadline() {
+        let deadline = Instant::now()
+            .checked_sub(Duration::from_millis(1))
+            .expect("past instant should be representable");
+        assert!(remaining_until(deadline).is_none());
+    }
+
+    #[tokio::test]
+    async fn test_run_with_lightpanda_deadline_enforces_timeout() {
+        let deadline = Instant::now() + Duration::from_millis(25);
+        let err = tokio::time::timeout(
+            Duration::from_secs(1),
+            run_with_lightpanda_deadline(
+                deadline,
+                async {
+                    sleep(Duration::from_millis(100)).await;
+                    Ok::<(), String>(())
+                },
+                "Target domain initialization attempt exceeded the remaining startup deadline",
+            ),
+        )
+        .await
+        .expect("outer timeout should not fire")
+        .unwrap_err();
+
+        assert!(err.contains(
+            "Timed out after 10000ms waiting for Lightpanda Target domain to initialize"
+        ));
+        assert!(err.contains("remaining startup deadline"));
+    }
+
+    #[tokio::test]
+    async fn test_run_with_lightpanda_deadline_returns_operation_error() {
+        let deadline = Instant::now() + Duration::from_secs(1);
+        let err = run_with_lightpanda_deadline(
+            deadline,
+            async { Err::<(), String>("Target.getTargets failed".to_string()) },
+            "unused timeout context",
+        )
+        .await
+        .unwrap_err();
+
+        assert_eq!(err, "Target.getTargets failed");
+    }
+
+    #[test]
+    fn test_lightpanda_target_init_timeout_includes_last_error() {
+        let err = lightpanda_target_init_timeout(Some("Target.setDiscoverTargets failed"));
+        assert!(err.contains(
+            "Timed out after 10000ms waiting for Lightpanda Target domain to initialize"
+        ));
+        assert!(err.contains("Target.setDiscoverTargets failed"));
+    }
 }

cli/src/native/cdp/chrome.rs

@@ -3,7 +3,7 @@ use std::path::{Path, PathBuf};
 use std::process::{Child, Command, Stdio};
 use std::time::Duration;
 
-use super::types::BrowserVersionInfo;
+use super::discovery::discover_cdp_url;
 
 pub struct ChromeProcess {
     child: Child,
@@ -386,28 +386,6 @@ pub fn find_chrome() -> Option<PathBuf> {
     None
 }
 
-pub async fn discover_cdp_url(port: u16) -> Result<String, String> {
-    let url = format!("http://127.0.0.1:{}/json/version", port);
-
-    let body = tokio::time::timeout(Duration::from_secs(2), async {
-        reqwest_get_string(&url).await
-    })
-    .await
-    .map_err(|_| format!("Timeout connecting to CDP on port {}", port))?
-    .map_err(|e| format!("Failed to connect to CDP on port {}: {}", port, e))?;
-
-    let info: BrowserVersionInfo = serde_json::from_str(&body)
-        .map_err(|e| format!("Invalid /json/version response: {}", e))?;
-
-    info.web_socket_debugger_url
-        .ok_or_else(|| format!("No webSocketDebuggerUrl in /json/version on port {}", port))
-}
-
-async fn reqwest_get_string(url: &str) -> Result<String, String> {
-    let resp = reqwest::get(url).await.map_err(|e| e.to_string())?;
-    resp.text().await.map_err(|e| e.to_string())
-}
-
 pub fn read_devtools_active_port(user_data_dir: &Path) -> Option<(u16, String)> {
     let path = user_data_dir.join("DevToolsActivePort");
     let content = std::fs::read_to_string(&path).ok()?;

cli/src/native/cdp/discovery.rs

@@ -0,0 +1,73 @@
+use std::time::Duration;
+
+use super::types::BrowserVersionInfo;
+
+pub async fn discover_cdp_url(port: u16) -> Result<String, String> {
+    discover_cdp_url_with_request_timeout(port, Duration::from_secs(2)).await
+}
+
+pub async fn discover_cdp_url_with_request_timeout(
+    port: u16,
+    request_timeout: Duration,
+) -> Result<String, String> {
+    let url = format!("http://127.0.0.1:{}/json/version", port);
+
+    let body = tokio::time::timeout(request_timeout, async { reqwest_get_string(&url).await })
+        .await
+        .map_err(|_| format!("Timeout connecting to CDP on port {}", port))?
+        .map_err(|e| format!("Failed to connect to CDP on port {}: {}", port, e))?;
+
+    let info: BrowserVersionInfo = serde_json::from_str(&body)
+        .map_err(|e| format!("Invalid /json/version response: {}", e))?;
+
+    info.web_socket_debugger_url
+        .ok_or_else(|| format!("No webSocketDebuggerUrl in /json/version on port {}", port))
+}
+
+async fn reqwest_get_string(url: &str) -> Result<String, String> {
+    let resp = reqwest::get(url).await.map_err(|e| e.to_string())?;
+    resp.text().await.map_err(|e| e.to_string())
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use tokio::io::{AsyncReadExt, AsyncWriteExt};
+    use tokio::net::TcpListener;
+
+    async fn spawn_json_server(body: &'static str) -> (u16, tokio::task::JoinHandle<()>) {
+        let listener = TcpListener::bind("127.0.0.1:0").await.unwrap();
+        let port = listener.local_addr().unwrap().port();
+        let handle = tokio::spawn(async move {
+            let (mut socket, _) = listener.accept().await.unwrap();
+            let mut buf = [0u8; 1024];
+            let _ = socket.read(&mut buf).await;
+            let response = format!(
+                "HTTP/1.1 200 OK\r\nContent-Length: {}\r\nConnection: close\r\nContent-Type: application/json\r\n\r\n{}",
+                body.len(),
+                body
+            );
+            socket.write_all(response.as_bytes()).await.unwrap();
+        });
+        (port, handle)
+    }
+
+    #[tokio::test]
+    async fn discovers_ws_url_from_json_version() {
+        let (port, server) =
+            spawn_json_server(r#"{"webSocketDebuggerUrl":"ws://127.0.0.1:1234/"}"#).await;
+
+        let ws_url = discover_cdp_url(port).await.unwrap();
+        assert_eq!(ws_url, "ws://127.0.0.1:1234/");
+        server.await.unwrap();
+    }
+
+    #[tokio::test]
+    async fn invalid_json_returns_parse_error() {
+        let (port, server) = spawn_json_server("not-json").await;
+
+        let err = discover_cdp_url(port).await.unwrap_err();
+        assert!(err.contains("Invalid /json/version response"));
+        server.await.unwrap();
+    }
+}

cli/src/native/cdp/lightpanda.rs

@@ -1,13 +1,23 @@
+use std::collections::VecDeque;
 use std::io::{BufRead, BufReader};
 use std::net::TcpListener;
 use std::path::PathBuf;
 use std::process::{Child, Command, Stdio};
+use std::sync::{Arc, Mutex};
 use std::time::Duration;
 
+use super::discovery::discover_cdp_url_with_request_timeout;
+
+const LIGHTPANDA_STARTUP_TIMEOUT: Duration = Duration::from_secs(10);
+const LIGHTPANDA_POLL_INTERVAL: Duration = Duration::from_millis(100);
+const LIGHTPANDA_DISCOVERY_TIMEOUT: Duration = Duration::from_millis(500);
+const LIGHTPANDA_SESSION_TIMEOUT_SECS: u64 = 604800; // 1 week, the documented maximum
+const MAX_LOG_LINES: usize = 40;
+
 pub struct LightpandaProcess {
     child: Child,
     pub ws_url: String,
-    _stderr_drain: Option<std::thread::JoinHandle<()>>,
+    _log_drainers: Vec<std::thread::JoinHandle<()>>,
 }
 
 impl LightpandaProcess {
@@ -30,8 +40,68 @@ pub struct LightpandaLaunchOptions {
     pub port: Option<u16>,
 }
 
+fn build_lightpanda_serve_args(port: u16, proxy: Option<&str>) -> Vec<String> {
+    let mut args = vec![
+        "serve".to_string(),
+        "--host".to_string(),
+        "127.0.0.1".to_string(),
+        "--port".to_string(),
+        port.to_string(),
+        "--timeout".to_string(),
+        LIGHTPANDA_SESSION_TIMEOUT_SECS.to_string(),
+    ];
+
+    if let Some(proxy) = proxy {
+        args.push("--http_proxy".to_string());
+        args.push(proxy.to_string());
+    }
+
+    args
+}
+
+#[derive(Clone, Default)]
+struct LaunchLogBuffer {
+    stdout: Arc<Mutex<VecDeque<String>>>,
+    stderr: Arc<Mutex<VecDeque<String>>>,
+}
+
+impl LaunchLogBuffer {
+    fn push_stdout(&self, line: String) {
+        push_bounded(&self.stdout, line);
+    }
+
+    fn push_stderr(&self, line: String) {
+        push_bounded(&self.stderr, line);
+    }
+
+    fn snapshot_stdout(&self) -> Vec<String> {
+        self.stdout
+            .lock()
+            .expect("stdout log buffer poisoned")
+            .iter()
+            .cloned()
+            .collect()
+    }
+
+    fn snapshot_stderr(&self) -> Vec<String> {
+        self.stderr
+            .lock()
+            .expect("stderr log buffer poisoned")
+            .iter()
+            .cloned()
+            .collect()
+    }
+}
+
+fn push_bounded(buffer: &Mutex<VecDeque<String>>, line: String) {
+    let mut guard = buffer.lock().expect("log buffer poisoned");
+    if guard.len() >= MAX_LOG_LINES {
+        guard.pop_front();
+    }
+    guard.push_back(line);
+}
+
 pub fn find_lightpanda() -> Option<PathBuf> {
-    // Check PATH via `which`
     #[cfg(unix)]
     {
         if let Ok(output) = Command::new("which").arg("lightpanda").output() {
@@ -61,7 +131,6 @@ pub fn find_lightpanda() -> Option<PathBuf> {
         }
     }
 
-    // Common install locations
     if let Some(home) = dirs::home_dir() {
         let candidates = [
             home.join(".lightpanda/lightpanda"),
@@ -74,13 +143,12 @@ pub fn find_lightpanda() -> Option<PathBuf> {
         }
     }
 
-    // npm package binary: @lightpanda/browser installs to node_modules/.bin
-    // Not checked here since the user would typically have it in PATH.
-
     None
 }
 
-pub fn launch_lightpanda(options: &LightpandaLaunchOptions) -> Result<LightpandaProcess, String> {
+pub async fn launch_lightpanda(
+    options: &LightpandaLaunchOptions,
+) -> Result<LightpandaProcess, String> {
     let binary_path = match &options.executable_path {
         Some(p) => PathBuf::from(p),
         None => find_lightpanda().ok_or(
@@ -95,31 +163,7 @@ pub fn launch_lightpanda(options: &LightpandaLaunchOptions) -> Result<Lightpanda
             .map(|a| a.port())
             .map_err(|e| format!("Failed to find an available port for Lightpanda: {}", e))?,
     };
-    let port_str = port.to_string();
-
-    let mut args = vec![
-        "serve".to_string(),
-        "--host".to_string(),
-        "127.0.0.1".to_string(),
-        "--port".to_string(),
-        port_str,
-    ];
-
-    if let Some(ref proxy) = options.proxy {
-        args.push("--http_proxy".to_string());
-        args.push(proxy.clone());
-    }
-
-    // Use the maximum inactivity timeout (1 week) so the connection stays alive
-    // during long sessions. Lightpanda treats 0 as "timeout immediately" rather
-    // than "no timeout".
-    args.push("--timeout".to_string());
-    args.push("604800".to_string());
-
-    // Lightpanda defaults to --log_level warn, which suppresses the "server
-    // running" info message we need to detect the listen address.
-    args.push("--log_level".to_string());
-    args.push("info".to_string());
+    let args = build_lightpanda_serve_args(port, options.proxy.as_deref());
 
     let mut child = Command::new(&binary_path)
         .args(&args)
@@ -129,174 +173,273 @@ pub fn launch_lightpanda(options: &LightpandaLaunchOptions) -> Result<Lightpanda
         .spawn()
         .map_err(|e| format!("Failed to launch Lightpanda at {:?}: {}", binary_path, e))?;
 
-    // Lightpanda logs to stderr
-    let stderr = child.stderr.take().ok_or_else(|| {
-        let _ = child.kill();
-        "Failed to capture Lightpanda stderr".to_string()
-    })?;
-    let reader = BufReader::new(stderr);
+    let (log_buffer, log_drainers) = start_log_drainers(&mut child)?;
 
-    let (address, reader) = match wait_for_address(reader) {
-        Ok(result) => result,
-        Err(e) => {
-            let _ = child.kill();
-            return Err(e);
-        }
-    };
-
-    let ws_url = format!("ws://{}", address);
-
-    let drain = std::thread::spawn(move || {
-        let mut reader = reader;
-        let mut buf = String::new();
-        loop {
-            buf.clear();
-            match reader.read_line(&mut buf) {
-                Ok(0) | Err(_) => break,
-                Ok(_) => {}
+    let ws_url =
+        match wait_for_lightpanda_ready(&mut child, port, &log_buffer, LIGHTPANDA_STARTUP_TIMEOUT)
+            .await
+        {
+            Ok(url) => url,
+            Err(e) => {
+                let _ = child.kill();
+                let _ = child.wait();
+                return Err(e);
             }
-        }
-    });
+        };
 
     Ok(LightpandaProcess {
         child,
         ws_url,
-        _stderr_drain: Some(drain),
+        _log_drainers: log_drainers,
     })
 }
 
-/// Parse Lightpanda's stderr for the server address.
-/// Lightpanda outputs lines like:
-///   INFO  app : server running . . . address = 127.0.0.1:9222
-///
-/// Returns the address and the reader so the caller can keep the pipe alive.
-fn wait_for_address(
-    mut reader: BufReader<std::process::ChildStderr>,
-) -> Result<(String, BufReader<std::process::ChildStderr>), String> {
-    let deadline = std::time::Instant::now() + Duration::from_secs(30);
-    let mut stderr_lines: Vec<String> = Vec::new();
-    let mut buf = String::new();
+fn start_log_drainers(
+    child: &mut Child,
+) -> Result<(LaunchLogBuffer, Vec<std::thread::JoinHandle<()>>), String> {
+    let stdout = child.stdout.take().ok_or_else(|| {
+        let _ = child.kill();
+        "Failed to capture Lightpanda stdout".to_string()
+    })?;
+    let stderr = child.stderr.take().ok_or_else(|| {
+        let _ = child.kill();
+        "Failed to capture Lightpanda stderr".to_string()
+    })?;
+
+    let logs = LaunchLogBuffer::default();
+    let stdout_logs = logs.clone();
+    let stderr_logs = logs.clone();
+
+    let stdout_handle =
+        std::thread::spawn(move || drain_reader(stdout, move |line| stdout_logs.push_stdout(line)));
+    let stderr_handle =
+        std::thread::spawn(move || drain_reader(stderr, move |line| stderr_logs.push_stderr(line)));
+
+    Ok((logs, vec![stdout_handle, stderr_handle]))
+}
+
+fn drain_reader<R, F>(reader: R, mut push: F)
+where
+    R: std::io::Read,
+    F: FnMut(String),
+{
+    for line in BufReader::new(reader).lines() {
+        match line {
+            Ok(line) => push(line),
+            Err(_) => break,
+        }
+    }
+}
+
+async fn wait_for_lightpanda_ready(
+    child: &mut Child,
+    port: u16,
+    logs: &LaunchLogBuffer,
+    startup_timeout: Duration,
+) -> Result<String, String> {
+    let deadline = std::time::Instant::now() + startup_timeout;
+    let mut last_probe_error = None;
 
     loop {
-        if std::time::Instant::now() > deadline {
+        if let Ok(Some(status)) = child.try_wait() {
+            // Give the drainer threads a brief window to flush the last log lines
+            // before we snapshot them.  This is best-effort: lines written just
+            // before exit may still be missing, but the most useful output (early
+            // startup errors) will already be in the buffer.
+            tokio::time::sleep(Duration::from_millis(25)).await;
             return Err(lightpanda_launch_error(
-                "Timeout waiting for Lightpanda server address",
-                &stderr_lines,
+                &format!(
+                    "Lightpanda exited before CDP became ready (status: {})",
+                    status
+                ),
+                logs,
+                last_probe_error.as_deref(),
             ));
         }
-        buf.clear();
-        match reader.read_line(&mut buf) {
-            Ok(0) => {
-                return Err(lightpanda_launch_error(
-                    "Lightpanda exited before providing server address",
-                    &stderr_lines,
-                ));
-            }
-            Ok(_) => {
-                let line = buf.trim_end().to_string();
-                if let Some(address) = extract_address(&line) {
-                    return Ok((address, reader));
-                }
-                stderr_lines.push(line);
-            }
-            Err(e) => {
-                return Err(format!("Failed to read Lightpanda stderr: {}", e));
-            }
+
+        match discover_cdp_url_with_request_timeout(port, LIGHTPANDA_DISCOVERY_TIMEOUT).await {
+            Ok(ws_url) => return Ok(ws_url),
+            Err(err) => last_probe_error = Some(err),
         }
+
+        if std::time::Instant::now() >= deadline {
+            return Err(lightpanda_launch_error(
+                &format!(
+                    "Timed out after {}ms waiting for Lightpanda CDP endpoint on port {}",
+                    startup_timeout.as_millis(),
+                    port
+                ),
+                logs,
+                last_probe_error.as_deref(),
+            ));
+        }
+
+        tokio::time::sleep(LIGHTPANDA_POLL_INTERVAL).await;
     }
 }
 
-fn extract_address(line: &str) -> Option<String> {
-    // Lightpanda uses logfmt (`address=...`) in release, pretty (`address = ...`) in debug.
-    for pattern in &["address=", "address = "] {
-        if let Some(idx) = line.find(pattern) {
-            let addr = line[idx + pattern.len()..].trim().to_string();
-            // logfmt lines may have subsequent key=value pairs
-            let addr = addr.split_whitespace().next().unwrap_or("").to_string();
-            if !addr.is_empty() {
-                return Some(addr);
-            }
-        }
-    }
-    None
-}
+fn lightpanda_launch_error(
+    message: &str,
+    logs: &LaunchLogBuffer,
+    last_probe_error: Option<&str>,
+) -> String {
+    let stdout_lines = logs.snapshot_stdout();
+    let stderr_lines = logs.snapshot_stderr();
+    let mut details = Vec::new();
 
-fn lightpanda_launch_error(message: &str, stderr_lines: &[String]) -> String {
-    if stderr_lines.is_empty() {
-        return format!("{} (no stderr output from Lightpanda)", message);
+    if let Some(err) = last_probe_error {
+        details.push(format!("Last probe error: {}", err));
     }
 
-    let last_lines: Vec<&String> = stderr_lines.iter().rev().take(5).collect();
-    format!(
-        "{}\nLightpanda stderr (last {} lines):\n  {}",
-        message,
-        last_lines.len(),
-        last_lines
-            .into_iter()
-            .rev()
-            .map(|s| s.as_str())
-            .collect::<Vec<_>>()
-            .join("\n  ")
-    )
+    if !stderr_lines.is_empty() {
+        details.push(format!(
+            "Lightpanda stderr (last {} lines):\n  {}",
+            stderr_lines.len(),
+            stderr_lines.join("\n  ")
+        ));
+    }
+
+    if !stdout_lines.is_empty() {
+        details.push(format!(
+            "Lightpanda stdout (last {} lines):\n  {}",
+            stdout_lines.len(),
+            stdout_lines.join("\n  ")
+        ));
+    }
+
+    if details.is_empty() {
+        format!("{} (no stdout/stderr output from Lightpanda)", message)
+    } else {
+        format!("{}\n{}", message, details.join("\n"))
+    }
 }
 
 #[cfg(test)]
 mod tests {
     use super::*;
+    use tokio::io::{AsyncReadExt, AsyncWriteExt};
+    use tokio::net::TcpListener as TokioTcpListener;
 
-    #[test]
-    fn test_extract_address_pretty_debug_build() {
-        assert_eq!(
-            extract_address("      address = 127.0.0.1:9222"),
-            Some("127.0.0.1:9222".to_string())
-        );
+    fn unused_port() -> u16 {
+        std::net::TcpListener::bind("127.0.0.1:0")
+            .unwrap()
+            .local_addr()
+            .unwrap()
+            .port()
     }
 
-    #[test]
-    fn test_extract_address_logfmt_release_build() {
-        assert_eq!(
-            extract_address(
-                "$time=1234 $scope=app $level=info $msg=\"server running\" address=127.0.0.1:9222"
-            ),
-            Some("127.0.0.1:9222".to_string())
+    async fn serve_json_version_once_after_delay(port: u16, delay_ms: u64, body: &'static str) {
+        tokio::time::sleep(Duration::from_millis(delay_ms)).await;
+        let listener = TokioTcpListener::bind(("127.0.0.1", port)).await.unwrap();
+        let (mut socket, _) = listener.accept().await.unwrap();
+        let mut buf = [0u8; 1024];
+        let _ = socket.read(&mut buf).await;
+        let response = format!(
+            "HTTP/1.1 200 OK\r\nContent-Length: {}\r\nConnection: close\r\nContent-Type: application/json\r\n\r\n{}",
+            body.len(),
+            body
         );
+        socket.write_all(response.as_bytes()).await.unwrap();
     }
 
-    #[test]
-    fn test_extract_address_pretty_inline() {
-        assert_eq!(
-            extract_address("INFO  app : server running address = 127.0.0.1:4567"),
-            Some("127.0.0.1:4567".to_string())
-        );
+    #[tokio::test]
+    async fn waits_for_ready_without_logs() {
+        let port = unused_port();
+        tokio::spawn(serve_json_version_once_after_delay(
+            port,
+            150,
+            r#"{"webSocketDebuggerUrl":"ws://127.0.0.1:9222/"}"#,
+        ));
+
+        let mut child = Command::new("/bin/sh")
+            .args(["-c", "sleep 5"])
+            .stdin(Stdio::null())
+            .stdout(Stdio::piped())
+            .stderr(Stdio::piped())
+            .spawn()
+            .unwrap();
+
+        let (logs, _drainers) = start_log_drainers(&mut child).unwrap();
+        let ws_url = wait_for_lightpanda_ready(&mut child, port, &logs, LIGHTPANDA_STARTUP_TIMEOUT)
+            .await
+            .unwrap();
+
+        assert_eq!(ws_url, "ws://127.0.0.1:9222/");
+        let _ = child.kill();
+        let _ = child.wait();
     }
 
-    #[test]
-    fn test_extract_address_no_match() {
-        assert_eq!(extract_address("INFO  app : starting up..."), None);
+    #[tokio::test]
+    async fn child_exit_surfaces_logs() {
+        let port = unused_port();
+        let mut child = Command::new("/bin/sh")
+            .args(["-c", "echo boom >&2; sleep 0.1; exit 23"])
+            .stdin(Stdio::null())
+            .stdout(Stdio::piped())
+            .stderr(Stdio::piped())
+            .spawn()
+            .unwrap();
+
+        let (logs, _drainers) = start_log_drainers(&mut child).unwrap();
+        let err = wait_for_lightpanda_ready(&mut child, port, &logs, LIGHTPANDA_STARTUP_TIMEOUT)
+            .await
+            .unwrap_err();
+
+        assert!(err.contains("Lightpanda exited before CDP became ready"));
+        assert!(err.contains("boom"));
+    }
+
+    #[tokio::test]
+    async fn timeout_reports_last_probe_error() {
+        let port = unused_port();
+        let mut child = Command::new("/bin/sh")
+            .args(["-c", "sleep 30"])
+            .stdin(Stdio::null())
+            .stdout(Stdio::piped())
+            .stderr(Stdio::piped())
+            .spawn()
+            .unwrap();
+
+        let timeout = Duration::from_millis(300);
+        let (logs, _drainers) = start_log_drainers(&mut child).unwrap();
+        let err = tokio::time::timeout(
+            Duration::from_secs(2),
+            wait_for_lightpanda_ready(&mut child, port, &logs, timeout),
+        )
+        .await
+        .expect("ready wait should return before outer timeout")
+        .unwrap_err();
+
+        assert!(err.contains("Timed out after 300ms waiting for Lightpanda CDP endpoint"));
+        assert!(
+            err.contains("Failed to connect to CDP") || err.contains("Timeout connecting to CDP")
+        );
+
+        let _ = child.kill();
+        let _ = child.wait();
     }
 
     #[test]
     fn test_find_lightpanda_returns_none_when_missing() {
-        // On most CI/dev machines Lightpanda won't be installed
-        // Just verify the function doesn't panic
         let _ = find_lightpanda();
     }
 
     #[test]
-    fn test_lightpanda_launch_error_no_stderr() {
-        let msg = lightpanda_launch_error("Lightpanda exited", &[]);
-        assert!(msg.contains("no stderr output"));
+    fn test_lightpanda_launch_error_no_logs() {
+        let logs = LaunchLogBuffer::default();
+        let msg = lightpanda_launch_error("Lightpanda exited", &logs, None);
+        assert!(msg.contains("no stdout/stderr output"));
     }
 
     #[test]
     fn test_lightpanda_launch_error_with_lines() {
-        let lines = vec![
-            "INFO starting up".to_string(),
-            "ERROR bind failed: address in use".to_string(),
-        ];
-        let msg = lightpanda_launch_error("Lightpanda exited", &lines);
-        assert!(msg.contains("bind failed"));
-        assert!(msg.contains("last 2 lines"));
+        let logs = LaunchLogBuffer::default();
+        logs.push_stdout("stdout line".to_string());
+        logs.push_stderr("stderr line".to_string());
+        let msg = lightpanda_launch_error("Lightpanda exited", &logs, Some("connect failed"));
+        assert!(msg.contains("stdout line"));
+        assert!(msg.contains("stderr line"));
+        assert!(msg.contains("Last probe error: connect failed"));
     }
 
     #[test]
@@ -306,4 +449,42 @@ mod tests {
         assert!(opts.proxy.is_none());
         assert!(opts.port.is_none());
     }
+
+    #[test]
+    fn test_build_lightpanda_serve_args_sets_explicit_session_timeout() {
+        let args = build_lightpanda_serve_args(9222, None);
+
+        assert_eq!(
+            args,
+            vec![
+                "serve".to_string(),
+                "--host".to_string(),
+                "127.0.0.1".to_string(),
+                "--port".to_string(),
+                "9222".to_string(),
+                "--timeout".to_string(),
+                "604800".to_string(),
+            ]
+        );
+    }
+
+    #[test]
+    fn test_build_lightpanda_serve_args_with_proxy() {
+        let args = build_lightpanda_serve_args(9333, Some("http://127.0.0.1:8080"));
+
+        assert_eq!(
+            args,
+            vec![
+                "serve".to_string(),
+                "--host".to_string(),
+                "127.0.0.1".to_string(),
+                "--port".to_string(),
+                "9333".to_string(),
+                "--timeout".to_string(),
+                "604800".to_string(),
+                "--http_proxy".to_string(),
+                "http://127.0.0.1:8080".to_string(),
+            ]
+        );
+    }
 }

cli/src/native/cdp/mod.rs

@@ -1,4 +1,5 @@
 pub mod chrome;
 pub mod client;
+pub mod discovery;
 pub mod lightpanda;
 pub mod types;

cli/src/native/e2e_tests.rs

@@ -86,6 +86,92 @@ async fn e2e_launch_navigate_evaluate_close() {
     assert_eq!(get_data(&resp)["closed"], true);
 }
 
+#[tokio::test]
+#[ignore]
+async fn e2e_lightpanda_launch_can_open_page() {
+    let lightpanda_bin = match std::env::var("LIGHTPANDA_BIN") {
+        Ok(path) if !path.is_empty() => path,
+        _ => return,
+    };
+
+    let mut state = DaemonState::new();
+
+    let resp = tokio::time::timeout(
+        tokio::time::Duration::from_secs(20),
+        execute_command(
+            &json!({
+                "id": "1",
+                "action": "launch",
+                "headless": true,
+                "engine": "lightpanda",
+                "executablePath": lightpanda_bin,
+            }),
+            &mut state,
+        ),
+    )
+    .await
+    .expect("Lightpanda launch should not hang");
+
+    assert_success(&resp);
+    assert_eq!(get_data(&resp)["launched"], true);
+
+    let resp = execute_command(
+        &json!({ "id": "2", "action": "navigate", "url": "https://example.com" }),
+        &mut state,
+    )
+    .await;
+    assert_success(&resp);
+    assert_eq!(get_data(&resp)["url"], "https://example.com/");
+    assert_eq!(get_data(&resp)["title"], "Example Domain");
+
+    let resp = execute_command(&json!({ "id": "3", "action": "close" }), &mut state).await;
+    assert_success(&resp);
+    assert_eq!(get_data(&resp)["closed"], true);
+}
+
+#[tokio::test]
+#[ignore]
+async fn e2e_lightpanda_auto_launch_can_open_page() {
+    let lightpanda_bin = match std::env::var("LIGHTPANDA_BIN") {
+        Ok(path) if !path.is_empty() => path,
+        _ => return,
+    };
+
+    let prev_engine = std::env::var("AGENT_BROWSER_ENGINE").ok();
+    let prev_path = std::env::var("AGENT_BROWSER_EXECUTABLE_PATH").ok();
+    std::env::set_var("AGENT_BROWSER_ENGINE", "lightpanda");
+    std::env::set_var("AGENT_BROWSER_EXECUTABLE_PATH", &lightpanda_bin);
+
+    let mut state = DaemonState::new();
+
+    let resp = tokio::time::timeout(
+        tokio::time::Duration::from_secs(20),
+        execute_command(
+            &json!({ "id": "1", "action": "navigate", "url": "https://example.com" }),
+            &mut state,
+        ),
+    )
+    .await
+    .expect("Lightpanda auto-launch should not hang");
+
+    match prev_engine {
+        Some(value) => std::env::set_var("AGENT_BROWSER_ENGINE", value),
+        None => std::env::remove_var("AGENT_BROWSER_ENGINE"),
+    }
+    match prev_path {
+        Some(value) => std::env::set_var("AGENT_BROWSER_EXECUTABLE_PATH", value),
+        None => std::env::remove_var("AGENT_BROWSER_EXECUTABLE_PATH"),
+    }
+
+    assert_success(&resp);
+    assert_eq!(get_data(&resp)["url"], "https://example.com/");
+    assert_eq!(get_data(&resp)["title"], "Example Domain");
+
+    let resp = execute_command(&json!({ "id": "2", "action": "close" }), &mut state).await;
+    assert_success(&resp);
+    assert_eq!(get_data(&resp)["closed"], true);
+}
+
 // ---------------------------------------------------------------------------
 // Snapshot with refs and ref-based click
 // ---------------------------------------------------------------------------