feat: auto-start stopped containers on boot, add failure recovery tests

Added start_stopped_containers() to crash_recovery.rs that starts all
exited/created containers on backend startup, fixing the issue where
containers didn't come back after clean reboot (PID marker removed by
systemd stop). Created test-failure-recovery.sh covering 5 failure
scenarios: container crash, backend restart, Tor restart, full reboot,
and Tor traffic block (UPTIME-02).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

core/archipelago/src/crash_recovery.rs

@@ -222,6 +222,37 @@ fn is_process_running(pid: u32) -> bool {
     std::path::Path::new(&format!("/proc/{}", pid)).exists()
 }
 
+/// Start all stopped containers that were previously installed.
+/// Runs on every startup to ensure containers come back after clean reboots.
+/// The crash recovery (PID-based) handles dirty shutdowns; this handles clean ones.
+pub async fn start_stopped_containers() -> RecoveryReport {
+    let output = tokio::process::Command::new("sudo")
+        .args(["podman", "ps", "-a", "--filter", "status=exited", "--filter", "status=created", "--format", "{{.Names}}"])
+        .output()
+        .await;
+
+    let names: Vec<String> = match output {
+        Ok(o) if o.status.success() => {
+            String::from_utf8_lossy(&o.stdout)
+                .lines()
+                .filter(|l| !l.is_empty())
+                .map(|s| s.to_string())
+                .collect()
+        }
+        _ => Vec::new(),
+    };
+
+    if names.is_empty() {
+        return RecoveryReport { total: 0, recovered: 0, failed: Vec::new() };
+    }
+
+    info!("Starting {} stopped containers after boot...", names.len());
+    let records: Vec<RunningContainerRecord> = names.iter()
+        .map(|n| RunningContainerRecord { name: n.clone(), image: String::new() })
+        .collect();
+    recover_containers(&records).await
+}
+
 /// Spawn a background task that periodically saves the container snapshot.
 pub fn spawn_snapshot_task(data_dir: PathBuf) {
     tokio::spawn(async move {

core/archipelago/src/main.rs

@@ -76,6 +76,15 @@ async fn main() -> Result<()> {
         );
     }
 
+    // Start any stopped containers (handles clean reboot where PID was removed)
+    let boot_report = crash_recovery::start_stopped_containers().await;
+    if boot_report.total > 0 {
+        info!(
+            "🔄 Boot startup: {}/{} containers started (failed: {:?})",
+            boot_report.recovered, boot_report.total, boot_report.failed
+        );
+    }
+
     // Write PID marker so we can detect crashes on next startup
     crash_recovery::write_pid_marker(&config.data_dir).await?;