refactor: rewrite tunnel system with host-network cloudflared containers

Replace the subprocess-based tunnel implementation with Docker containers
running on the host network. This eliminates all container name resolution
bugs that caused tunnel 502 errors.

New design:
- Each tunnel is a docker run --network host cloudflare/cloudflared container
- cloudflared connects to localhost:{published_port} (Docker port forwarding)
- No dependency on container names, backend network DNS, or binding diagnostics
- Tunnels named predictably: tunnel-{instance_name}
- Start/stop/recreate use container names instead of PIDs

Files changed:
- NEW: apps/api/src/services/tunnel.py — clean tunnel module (start/stop/recreate/health)
- apps/api/src/services/docker.py — removed 250 lines of old tunnel code
- apps/api/src/api/tool_instances.py — use new tunnel module, store container_name
- apps/api/src/services/health_monitor.py — updated import
- apps/web/src/components/session-card.tsx — Recreate Tunnel button always visible

Quality gates: ruff clean, 13 tests passed (health_monitor + notifications)

apps/api/src/api/tool_instances.py

@@ -45,7 +45,6 @@ from src.services.config_profile_resolver import (
     resolve_profile,
 )
 from src.services.docker import (
-    check_tunnel_health,
     connect_container_to_network,
     ensure_instance_directory,
     execute_compose_command,
@@ -53,16 +52,19 @@ from src.services.docker import (
     get_container_id,
     get_container_logs,
     get_container_status,
-    recreate_tunnel,
     render_compose_template,
     sort_volumes_by_specificity,
-    start_cloudflared_tunnel,
-    stop_cloudflared_tunnel,
     wait_for_container_running,
     write_compose_file,
     write_config_files,
     write_env_file,
 )
+from src.services.tunnel import (
+    check_tunnel_health,
+    recreate_tunnel,
+    start_tunnel,
+    stop_tunnel,
+)
 from src.services.docker_build import build_image
 from src.services.manifest_compiler import (
     compile_compose,
@@ -1960,12 +1962,12 @@ async def start_instance(
             "error": f"Tool type '{instance.tool_type_id}' not found",
         }
 
-    instance_port = tool_type.default_port or 0
+    published_port = instance.port or 0
     logger.debug(
-        "Tool type for instance %s: name=%s, default_port=%s, interface_type=%s",
+        "Tool type for instance %s: name=%s, published_port=%s, interface_type=%s",
         instance.id,
         tool_type.name,
-        instance_port,
+        published_port,
         tool_type.interface_type,
     )
 
@@ -1974,23 +1976,22 @@ async def start_instance(
         # Create temporary Cloudflare tunnel for public access
         try:
             logger.debug(
-                "Creating temporary tunnel for instance %s (container=%s, port=%d)",
+                "Creating tunnel for instance %s (port=%d)",
                 instance.id,
-                instance.container_name,
+                published_port,
-                instance_port,
             )
-            tunnel_info = start_cloudflared_tunnel(
+            tunnel_info = start_tunnel(
-                container_name=instance.container_name or instance.name,
+                instance_name=instance.name,
-                port=instance_port,
+                published_port=published_port,
             )
-            instance.tunnel_id = tunnel_info["pid"]
+            instance.tunnel_id = tunnel_info["container_name"]
             instance.public_url = tunnel_info["url"]
             instance.url = tunnel_info["url"]
             await session.commit()
             logger.debug(
-                "Created temporary tunnel for instance %s: pid=%s, url=%s",
+                "Created tunnel for instance %s: container=%s, url=%s",
                 instance.id,
-                tunnel_info["pid"],
+                tunnel_info["container_name"],
                 tunnel_info["url"],
             )
         except Exception as exc:
@@ -2060,9 +2061,9 @@ async def stop_instance(
     # Stop Cloudflare tunnel if exists
     if instance.tunnel_id:
         try:
-            stop_cloudflared_tunnel(instance.tunnel_id)
+            stop_tunnel(instance.name)
             logger.debug(
-                "Stopped tunnel for instance %s (pid=%s)",
+                "Stopped tunnel for instance %s (container=%s)",
                 instance.id,
                 instance.tunnel_id,
             )
@@ -2129,9 +2130,9 @@ async def restart_instance(
     # Stop old tunnel if exists
     if instance.tunnel_id:
         try:
-            stop_cloudflared_tunnel(instance.tunnel_id)
+            stop_tunnel(instance.name)
             logger.debug(
-                "Stopped old tunnel for instance %s (pid=%s)",
+                "Stopped old tunnel for instance %s (container=%s)",
                 instance.id,
                 instance.tunnel_id,
             )
@@ -2197,17 +2198,15 @@ async def restart_instance(
                     "error": f"Tool type '{tool_type.name if tool_type else 'unknown'}' has no port configured",
                 }
 
-            instance_port = tool_type.default_port
-
             # Only create tunnel for web-enabled tools
             if tool_type.interface_type == "web":
-                # Create new temporary tunnel
+                # Create new tunnel
                 try:
-                    tunnel_info = start_cloudflared_tunnel(
+                    tunnel_info = start_tunnel(
-                        container_name=instance.name.lower(),
+                        instance_name=instance.name,
-                        port=instance_port,
+                        published_port=instance.port or 0,
                     )
-                    instance.tunnel_id = tunnel_info["pid"]
+                    instance.tunnel_id = tunnel_info["container_name"]
                     instance.public_url = tunnel_info["url"]
                     instance.url = tunnel_info["url"]
                     logger.debug(
@@ -2306,9 +2305,9 @@ async def delete_instance(
     # Stop Cloudflare tunnel if exists
     if instance.tunnel_id:
         try:
-            stop_cloudflared_tunnel(instance.tunnel_id)
+            stop_tunnel(instance.name)
             logger.debug(
-                "Stopped tunnel for instance %s (pid=%s)",
+                "Stopped tunnel for instance %s (container=%s)",
                 instance.id,
                 instance.tunnel_id,
             )
@@ -2438,26 +2437,19 @@ async def recreate_tunnel_endpoint(
                 "message": "Tunnel is already healthy",
             }
 
-    # Get tool type for default port
-    tool_type = await session.get(ToolType, instance.tool_type_id)
-    instance_port = (
-        tool_type.default_port if tool_type and tool_type.default_port else 8080
-    )
-
     try:
         tunnel_info = recreate_tunnel(
-            container_name=instance.container_name or instance.name,
+            instance_name=instance.name,
-            port=instance_port,
+            published_port=instance.port or 0,
-            old_pid=instance.tunnel_id,
         )
-        instance.tunnel_id = tunnel_info["pid"]
+        instance.tunnel_id = tunnel_info["container_name"]
         instance.public_url = tunnel_info["url"]
         instance.url = tunnel_info["url"]
         await session.commit()
         logger.debug(
-            "Recreated tunnel for instance %s: pid=%s, url=%s",
+            "Recreated tunnel for instance %s: container=%s, url=%s",
             instance.id,
-            tunnel_info["pid"],
+            tunnel_info["container_name"],
             tunnel_info["url"],
         )
         return {"status": "healthy", "url": instance.url}

apps/api/src/services/docker.py

@@ -1,8 +1,6 @@
 """Docker service for managing tool instances."""
 
 import logging
-import os
-import re
 import subprocess
 import time
 from collections import Counter
@@ -384,334 +382,4 @@ def find_free_port(start: int = 10000, end: int = 20000) -> int:
     raise RuntimeError(f"No free port found in range {start}-{end}")
 
 
-def _check_app_binding(container_name: str, port: int) -> dict[str, str | bool]:
-    """Diagnose whether the app is bound to 127.0.0.1 or 0.0.0.0.
 
-    Checks from both inside the container (localhost) and outside
-    (via Docker network) to detect binding issues.
-
-    Returns:
-        Dict with 'internal_ok', 'external_ok', 'internal_status',
-        'external_status', and 'diagnosis'.
-    """
-    import subprocess
-
-    result: dict[str, Any] = {
-        "internal_ok": False,
-        "external_ok": False,
-        "internal_status": None,
-        "external_status": None,
-        "diagnosis": "unknown",
-    }
-
-    # Check from inside the container (loopback)
-    internal = subprocess.run(
-        [
-            "docker",
-            "exec",
-            container_name,
-            "sh",
-            "-c",
-            f"curl -s -o /dev/null -w '%{{http_code}}' http://localhost:{port}",
-        ],
-        capture_output=True,
-        text=True,
-        timeout=5,
-    )
-    if internal.returncode == 0:
-        try:
-            result["internal_status"] = int(internal.stdout.strip())
-            result["internal_ok"] = result["internal_status"] > 0
-        except ValueError:
-            pass
-
-    # Check from outside the container (Docker network)
-    external = subprocess.run(
-        [
-            "curl",
-            "-s",
-            "-o",
-            "/dev/null",
-            "-w",
-            "%{http_code}",
-            f"http://{container_name}:{port}",
-        ],
-        capture_output=True,
-        text=True,
-        timeout=5,
-    )
-    if external.returncode == 0:
-        try:
-            result["external_status"] = int(external.stdout.strip())
-            result["external_ok"] = result["external_status"] > 0
-        except ValueError:
-            pass
-
-    # Diagnose binding issue
-    if result["internal_ok"] and not result["external_ok"]:
-        result["diagnosis"] = (
-            f"App appears to be bound to 127.0.0.1:{port} inside the container. "
-            f"It must bind to 0.0.0.0:{port} to be accessible from the tunnel."
-        )
-    elif result["internal_ok"] and result["external_ok"]:
-        result["diagnosis"] = "App is accessible on both interfaces."
-    elif not result["internal_ok"] and not result["external_ok"]:
-        result["diagnosis"] = f"App is not responding on port {port} at all."
-    else:
-        result["diagnosis"] = "Unexpected binding state."
-
-    return result
-
-
-def start_cloudflared_tunnel(
-    container_name: str, port: int, timeout: int = 30
-) -> dict[str, str]:
-    """Start a temporary Cloudflare tunnel for a container.
-
-    Uses 'cloudflared tunnel --url' to create a temporary tunnel
-    with a random trycloudflare.com URL.
-
-    Args:
-        container_name: Name of the Docker container to tunnel to
-        port: Port number the container listens on
-        timeout: Maximum seconds to wait for tunnel URL
-
-    Returns:
-        Dict with 'url' (the public tunnel URL) and 'pid' (process ID)
-    """
-    import subprocess
-    import logging
-
-    logger = logging.getLogger(__name__)
-
-    # First verify the container is accessible from the Docker network
-    logger.info("Checking connectivity to %s:%d...", container_name, port)
-    accessible = False
-    last_status = None
-    for attempt in range(30):  # 30 attempts × 1s = 30s max wait for app startup
-        check = subprocess.run(
-            [
-                "curl",
-                "-s",
-                "-o",
-                "/dev/null",
-                "-w",
-                "%{http_code}",
-                "--max-time",
-                "3",
-                f"http://{container_name}:{port}",
-            ],
-            capture_output=True,
-            text=True,
-            timeout=5,
-        )
-        status_str = check.stdout.strip()
-        logger.info(
-            "Connectivity check %d/%d: http_code=%s (rc=%d)",
-            attempt + 1,
-            30,
-            status_str,
-            check.returncode,
-        )
-        try:
-            last_status = int(status_str)
-            # Accept 2xx, 3xx, 401, 403 as "app is listening"
-            if last_status in (401, 403) or 200 <= last_status < 400:
-                accessible = True
-                logger.info(
-                    "App on %s:%d is ready (HTTP %d)",
-                    container_name,
-                    port,
-                    last_status,
-                )
-                break
-        except ValueError:
-            pass
-
-        if check.returncode != 0:
-            logger.debug(
-                "curl failed: stderr=%s", check.stderr.strip() if check.stderr else ""
-            )
-        time.sleep(1)
-
-    if not accessible:
-        logger.warning(
-            "Container %s:%d not responding after 30s (last status: %s). "
-            "Running binding diagnostics...",
-            container_name,
-            port,
-            last_status,
-        )
-        diagnosis = _check_app_binding(container_name, port)
-        logger.warning(
-            "Binding diagnosis: internal=%s (HTTP %s), external=%s (HTTP %s). %s",
-            diagnosis["internal_ok"],
-            diagnosis["internal_status"],
-            diagnosis["external_ok"],
-            diagnosis["external_status"],
-            diagnosis["diagnosis"],
-        )
-
-    # Run cloudflared in background, capture output
-    logger.info("Starting cloudflared tunnel to http://%s:%d", container_name, port)
-    proc = subprocess.Popen(
-        ["cloudflared", "tunnel", "--url", f"http://{container_name}:{port}"],
-        stdout=subprocess.PIPE,
-        stderr=subprocess.STDOUT,
-        text=True,
-    )
-
-    # Wait for the URL to appear in output
-    url_pattern = re.compile(r"https://[a-z0-9-]+\.trycloudflare\.com")
-    start_time = time.time()
-    url = None
-
-    if proc.stdout is None:
-        proc.terminate()
-        proc.wait(timeout=5)
-        raise RuntimeError("Failed to capture cloudflared output")
-
-    while time.time() - start_time < timeout:
-        # Read available output
-        import select
-
-        readable, _, _ = select.select([proc.stdout], [], [], 1.0)
-        if readable:
-            line = proc.stdout.readline()
-            if line:
-                match = url_pattern.search(line)
-                if match:
-                    url = match.group(0)
-                    break
-
-    if not url:
-        proc.terminate()
-        proc.wait(timeout=5)
-        raise RuntimeError(
-            f"Failed to get tunnel URL within {timeout}s. "
-            f"cloudflared output may contain errors."
-        )
-
-    return {"url": url, "pid": str(proc.pid)}
-
-
-def stop_cloudflared_tunnel(pid: str) -> None:
-    """Stop a cloudflared tunnel process.
-
-    Args:
-        pid: Process ID of the cloudflared tunnel
-    """
-    import signal
-
-    try:
-        os.kill(int(pid), signal.SIGTERM)
-    except ProcessLookupError:
-        pass  # Already stopped
-
-
-def recreate_tunnel(
-    container_name: str, port: int, old_pid: str | None = None
-) -> dict[str, str]:
-    """Recreate a temporary Cloudflare tunnel.
-
-    Stops the old tunnel (if pid provided) and starts a new one.
-
-    Args:
-        container_name: Name of the Docker container to tunnel to
-        port: Port number the container listens on
-        old_pid: Optional PID of the old tunnel process to stop
-
-    Returns:
-        Dict with 'url' and 'pid' for the new tunnel
-    """
-    if old_pid:
-        stop_cloudflared_tunnel(old_pid)
-
-    return start_cloudflared_tunnel(container_name, port)
-
-
-def check_tunnel_health(url: str, timeout: int = 10) -> dict[str, Any]:
-    """Check if a tunnel URL is healthy with smart error classification.
-
-    Args:
-        url: The tunnel URL to check
-        timeout: Request timeout in seconds
-
-    Returns:
-        Dict with 'tunnel_status' (healthy, unreachable, error_response, not_applicable),
-        'status_code' (int or None), 'healthy' (bool), and 'error' (str or None)
-    """
-    import subprocess
-
-    try:
-        result = subprocess.run(
-            [
-                "curl",
-                "-s",
-                "-o",
-                "/dev/null",
-                "-w",
-                "%{http_code}",
-                "--max-time",
-                str(timeout),
-                url,
-            ],
-            capture_output=True,
-            text=True,
-            timeout=timeout + 5,
-        )
-        status_code = int(result.stdout.strip())
-
-        if 200 <= status_code < 400:
-            return {
-                "tunnel_status": "healthy",
-                "status_code": status_code,
-                "healthy": True,
-                "error": None,
-            }
-        elif status_code in (502, 503, 504):
-            # Application error, not tunnel error
-            return {
-                "tunnel_status": "error_response",
-                "status_code": status_code,
-                "healthy": False,
-                "error": f"Application returned HTTP {status_code}",
-            }
-        else:
-            return {
-                "tunnel_status": "error_response",
-                "status_code": status_code,
-                "healthy": False,
-                "error": f"HTTP {status_code}",
-            }
-    except subprocess.TimeoutExpired:
-        return {
-            "tunnel_status": "unreachable",
-            "status_code": None,
-            "healthy": False,
-            "error": "Tunnel request timed out",
-        }
-    except (ValueError, Exception) as e:
-        error_str = str(e).lower()
-        # Classify connection errors
-        if any(
-            err in error_str
-            for err in [
-                "connection refused",
-                "econnrefused",
-                "could not resolve",
-                "nodename",
-            ]
-        ):
-            return {
-                "tunnel_status": "unreachable",
-                "status_code": None,
-                "healthy": False,
-                "error": f"Tunnel unreachable: {e}",
-            }
-        return {
-            "tunnel_status": "unreachable",
-            "status_code": None,
-            "healthy": False,
-            "error": str(e),
-        }

apps/api/src/services/health_monitor.py

@@ -13,7 +13,8 @@ from src.database import SessionLocal
 from src.models.health_check import HealthCheck
 from src.models.tool_instance import ToolInstance
 from src.services.correlation import get_correlation_id
-from src.services.docker import check_tunnel_health, get_container_status
+from src.services.docker import get_container_status
+from src.services.tunnel import check_tunnel_health
 from src.services.event_bus import InstanceEventBus, InstanceEventPayload
 from src.services.notification_service import notification_service
 

apps/api/src/services/tunnel.py

@@ -0,0 +1,213 @@
+"""Clean tunnel service using host-network cloudflared containers.
+
+Design:
+- Each tunnel runs as a Docker container on the host network.
+- cloudflared connects to localhost:{published_port}, leveraging Docker's
+  port forwarding. No container name resolution is required.
+- Tunnels are named predictably (tunnel-{instance_name}) for start/stop.
+"""
+
+import logging
+import re
+import subprocess
+from typing import Any
+
+logger = logging.getLogger(__name__)
+
+TUNNEL_IMAGE = "cloudflare/cloudflared:latest"
+
+
+def _tunnel_container_name(instance_name: str) -> str:
+    return f"tunnel-{instance_name.lower()}"
+
+
+def _ensure_image() -> None:
+    """Pull cloudflared image if not already present."""
+    result = subprocess.run(
+        ["docker", "images", "-q", TUNNEL_IMAGE],
+        capture_output=True,
+        text=True,
+    )
+    if not result.stdout.strip():
+        logger.info("Pulling %s ...", TUNNEL_IMAGE)
+        pull = subprocess.run(
+            ["docker", "pull", TUNNEL_IMAGE],
+            capture_output=True,
+            text=True,
+        )
+        if pull.returncode != 0:
+            logger.warning("Failed to pull %s: %s", TUNNEL_IMAGE, pull.stderr)
+
+
+def start_tunnel(instance_name: str, published_port: int, timeout: int = 30) -> dict[str, str]:
+    """Start a temporary Cloudflare tunnel for an instance.
+
+    Args:
+        instance_name: The tool instance name (used to derive tunnel container name).
+        published_port: The host port Docker forwards to the container.
+        timeout: Seconds to wait for the tunnel URL.
+
+    Returns:
+        Dict with 'url' and 'container_name'.
+    """
+    _ensure_image()
+
+    tunnel_name = _tunnel_container_name(instance_name)
+
+    # Clean up any stale tunnel container with this name
+    subprocess.run(
+        ["docker", "stop", "-t", "3", tunnel_name],
+        capture_output=True,
+        text=True,
+    )
+    subprocess.run(
+        ["docker", "rm", tunnel_name],
+        capture_output=True,
+        text=True,
+    )
+
+    cmd = [
+        "docker", "run", "-d", "--rm",
+        "--network", "host",
+        "--name", tunnel_name,
+        TUNNEL_IMAGE,
+        "tunnel", "--url", f"http://localhost:{published_port}",
+    ]
+
+    proc = subprocess.run(cmd, capture_output=True, text=True)
+    if proc.returncode != 0:
+        raise RuntimeError(
+            f"Failed to start tunnel container {tunnel_name}: {proc.stderr}"
+        )
+
+    # Wait for URL to appear in logs
+    url_pattern = re.compile(r"https://[a-z0-9-]+\.trycloudflare\.com")
+    start_time = __import__("time").time()
+    url: str | None = None
+    logs_result = subprocess.run(
+        ["docker", "logs", tunnel_name],
+        capture_output=True,
+        text=True,
+    )
+
+    while __import__("time").time() - start_time < timeout:
+        match = url_pattern.search(logs_result.stdout)
+        if match:
+            url = match.group(0)
+            break
+        __import__("time").sleep(0.5)
+        logs_result = subprocess.run(
+            ["docker", "logs", tunnel_name],
+            capture_output=True,
+            text=True,
+        )
+
+    if not url:
+        subprocess.run(["docker", "stop", "-t", "3", tunnel_name], capture_output=True)
+        raise RuntimeError(
+            f"Tunnel {tunnel_name} did not produce a URL within {timeout}s. "
+            f"cloudflared logs:\n{logs_result.stdout[-2000:]}"
+        )
+
+    logger.info(
+        "Tunnel %s started for %s on port %d → %s",
+        tunnel_name,
+        instance_name,
+        published_port,
+        url,
+    )
+    return {"url": url, "container_name": tunnel_name}
+
+
+def stop_tunnel(instance_name: str) -> None:
+    """Stop the tunnel container for an instance."""
+    tunnel_name = _tunnel_container_name(instance_name)
+    subprocess.run(
+        ["docker", "stop", "-t", "5", tunnel_name],
+        capture_output=True,
+        text=True,
+    )
+    logger.debug("Stopped tunnel container %s", tunnel_name)
+
+
+def recreate_tunnel(instance_name: str, published_port: int) -> dict[str, str]:
+    """Recreate a tunnel for an instance."""
+    stop_tunnel(instance_name)
+    return start_tunnel(instance_name, published_port)
+
+
+def check_tunnel_health(url: str, timeout: int = 10) -> dict[str, Any]:
+    """Check if a tunnel URL is healthy.
+
+    Returns:
+        Dict with 'tunnel_status', 'status_code', 'healthy', 'error'.
+    """
+    try:
+        result = subprocess.run(
+            [
+                "curl",
+                "-s",
+                "-o",
+                "/dev/null",
+                "-w",
+                "%{http_code}",
+                "--max-time",
+                str(timeout),
+                url,
+            ],
+            capture_output=True,
+            text=True,
+            timeout=timeout + 5,
+        )
+        status_code = int(result.stdout.strip())
+
+        if 200 <= status_code < 400:
+            return {
+                "tunnel_status": "healthy",
+                "status_code": status_code,
+                "healthy": True,
+                "error": None,
+            }
+        if status_code in (502, 503, 504):
+            return {
+                "tunnel_status": "error_response",
+                "status_code": status_code,
+                "healthy": False,
+                "error": f"Application returned HTTP {status_code}",
+            }
+        return {
+            "tunnel_status": "error_response",
+            "status_code": status_code,
+            "healthy": False,
+            "error": f"HTTP {status_code}",
+        }
+    except subprocess.TimeoutExpired:
+        return {
+            "tunnel_status": "unreachable",
+            "status_code": None,
+            "healthy": False,
+            "error": "Tunnel request timed out",
+        }
+    except (ValueError, Exception) as exc:
+        error_str = str(exc).lower()
+        if any(
+            err in error_str
+            for err in [
+                "connection refused",
+                "econnrefused",
+                "could not resolve",
+                "nodename",
+            ]
+        ):
+            return {
+                "tunnel_status": "unreachable",
+                "status_code": None,
+                "healthy": False,
+                "error": f"Tunnel unreachable: {exc}",
+            }
+        return {
+            "tunnel_status": "unreachable",
+            "status_code": None,
+            "healthy": False,
+            "error": str(exc),
+        }

apps/web/src/components/session-card.tsx

@@ -333,7 +333,7 @@ export function SessionCard({
 									onClick: () => onRecreateTunnel(session),
 								},
 							]
-							: []),
+						: []),
 					...(isActive && onStop
 						? [
 								{