chore(nix): fix hash for updated go dependencies

Bumps [actions/stale](https://github.com/actions/stale) from 10.1.1 to 10.2.0.
- [Release notes](https://github.com/actions/stale/releases)
- [Changelog](https://github.com/actions/stale/blob/main/CHANGELOG.md)
- [Commits](997185467f...b5d41d4e1d)

---
updated-dependencies:
- dependency-name: actions/stale
  dependency-version: 10.2.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>

.env.example

@@ -1,5 +1,5 @@
 # Copy this file to .env and fill in your values
 # Required for connecting to Pangolin service
-PANGOLIN_ENDPOINT=https://example.com
+PANGOLIN_ENDPOINT=https://app.pangolin.net
 NEWT_ID=changeme-id
 NEWT_SECRET=changeme-secret

.github/workflows/nix-build.yml

@@ -0,0 +1,23 @@
+name: Build Nix package
+
+on:
+  workflow_dispatch:
+  pull_request:
+    paths:
+      - go.mod
+      - go.sum
+
+jobs:
+  nix-build:
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v6
+
+      - name: Install Nix
+        uses: DeterminateSystems/nix-installer-action@main
+
+      - name: Build flake package
+        run: |
+          nix build .#pangolin-newt -L

.github/workflows/nix-dependabot-update-hash.yml

@@ -0,0 +1,48 @@
+name: Update Nix Package Hash On Dependabot PRs
+
+on:
+  pull_request:
+    types: [opened, synchronize]
+    branches:
+      - main
+
+jobs:
+  nix-update:
+    if: github.actor == 'dependabot[bot]'
+    runs-on: ubuntu-latest
+
+    permissions:
+      contents: write
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v6
+        with:
+          ref: ${{ github.head_ref }}
+          token: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Install Nix
+        uses: DeterminateSystems/nix-installer-action@main
+
+      - name: Run nix-update
+        run: |
+          nix run nixpkgs#nix-update -- --flake pangolin-newt --no-src --version skip
+
+      - name: Check for changes
+        id: changes
+        run: |
+          if git diff --quiet; then
+            echo "changed=false" >> "$GITHUB_OUTPUT"
+          else
+            echo "changed=true" >> "$GITHUB_OUTPUT"
+          fi
+
+      - name: Commit and push changes
+        if: steps.changes.outputs.changed == 'true'
+        run: |
+          git config user.name "dependabot[bot]"
+          git config user.email "dependabot[bot]@users.noreply.github.com"
+
+          git add .
+          git commit -m "chore(nix): fix hash for updated go dependencies"
+          git push

.github/workflows/stale-bot.yml

@@ -0,0 +1,37 @@
+name: Mark and Close Stale Issues
+
+on:
+  schedule:
+    - cron: '0 0 * * *'
+  workflow_dispatch: # Allow manual trigger
+
+permissions:
+  contents: write # only for delete-branch option
+  issues: write
+  pull-requests: write
+
+jobs:
+  stale:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/stale@b5d41d4e1d5dceea10e7104786b73624c18a190f # v10.2.0
+        with:
+          days-before-stale: 14
+          days-before-close: 14
+          stale-issue-message: 'This issue has been automatically marked as stale due to 14 days of inactivity. It will be closed in 14 days if no further activity occurs.'
+          close-issue-message: 'This issue has been automatically closed due to inactivity. If you believe this is still relevant, please open a new issue with up-to-date information.'
+          stale-issue-label: 'stale'
+          
+          exempt-issue-labels: 'needs investigating, networking, new feature, reverse proxy, bug, api, authentication, documentation, enhancement, help wanted, good first issue, question'
+          
+          exempt-all-issue-assignees: true
+          
+          only-labels: ''
+          exempt-pr-labels: ''
+          days-before-pr-stale: -1
+          days-before-pr-close: -1
+          
+          operations-per-run: 100
+          remove-stale-when-updated: true
+          delete-branch: false
+          enable-statistics: true

.github/workflows/test.yml

@@ -10,22 +10,30 @@ on:
       - dev
 
 jobs:
-  test:
-    runs-on: amd64-runner
-
+  build:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        target:
+          - local
+          - docker-build
+          - go-build-release-darwin-amd64
+          - go-build-release-darwin-arm64
+          - go-build-release-freebsd-amd64
+          - go-build-release-freebsd-arm64
+          - go-build-release-linux-amd64
+          - go-build-release-linux-arm32-v6
+          - go-build-release-linux-arm32-v7
+          - go-build-release-linux-riscv64
+          - go-build-release-windows-amd64
     steps:
-      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
+      - name: Checkout repository
+        uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8 # v6.0.1
 
       - name: Set up Go
-        uses: actions/setup-go@44694675825211faa026b3c33043df3e48a5fa00 # v6.0.0
+        uses: actions/setup-go@7a3fe6cf4cb3a834922a1244abfce67bcef6a0c5 # v6.2.0
         with:
           go-version: 1.25
 
-      - name: Build go
-        run: go build
-
-      - name: Build Docker image
-        run: make build
-
-      - name: Build binaries
-        run: make go-build-release
+      - name: Build targets via `make`
+        run: make ${{ matrix.target }}

.gitignore

@@ -1,4 +1,3 @@
-newt
 .DS_Store
 bin/
 nohup.out
@@ -6,4 +5,6 @@ nohup.out
 *.iml
 certs/
 newt_arm64
-key
+key
+/.direnv/
+/result*

Dockerfile

@@ -1,4 +1,5 @@
-FROM golang:1.25-alpine AS builder
+# FROM golang:1.25-alpine AS builder
+FROM public.ecr.aws/docker/library/golang:1.25-alpine AS builder
 
 # Install git and ca-certificates
 RUN apk --no-cache add ca-certificates git tzdata
@@ -16,11 +17,12 @@ RUN go mod download
 COPY . .
 
 # Build the application
-RUN CGO_ENABLED=0 GOOS=linux go build -ldflags="-s -w" -o /newt
+ARG VERSION=dev
+RUN CGO_ENABLED=0 GOOS=linux go build -ldflags="-s -w -X main.newtVersion=${VERSION}" -o /newt
 
-FROM alpine:3.22 AS runner
+FROM public.ecr.aws/docker/library/alpine:3.23 AS runner
 
-RUN apk --no-cache add ca-certificates tzdata
+RUN apk --no-cache add ca-certificates tzdata iputils
 
 COPY --from=builder /newt /usr/local/bin/
 COPY entrypoint.sh /

Makefile

@@ -1,37 +1,73 @@
+.PHONY: all local docker-build docker-build-release
 
-all: build push
+all: local
+
+VERSION ?= dev
+LDFLAGS = -X main.newtVersion=$(VERSION)
+
+local:
+	CGO_ENABLED=0 go build -o ./bin/newt
+
+docker-build:
+	docker build -t fosrl/newt:latest .
 
 docker-build-release:
 	@if [ -z "$(tag)" ]; then \
 		echo "Error: tag is required. Usage: make docker-build-release tag=<tag>"; \
 		exit 1; \
 	fi
-	docker buildx build --platform linux/arm/v7,linux/arm64,linux/amd64 -t fosrl/newt:latest -f Dockerfile --push .
-	docker buildx build --platform linux/arm/v7,linux/arm64,linux/amd64 -t fosrl/newt:$(tag) -f Dockerfile --push .
+	docker buildx build . \
+		--platform linux/arm/v7,linux/arm64,linux/amd64 \
+		-t fosrl/newt:latest \
+		-t fosrl/newt:$(tag) \
+		-f Dockerfile \
+		--push
 
-build:
-	docker build -t fosrl/newt:latest .
+.PHONY: go-build-release \
+        go-build-release-linux-arm64 go-build-release-linux-arm32-v7 \
+        go-build-release-linux-arm32-v6 go-build-release-linux-amd64 \
+        go-build-release-linux-riscv64 go-build-release-darwin-arm64 \
+        go-build-release-darwin-amd64 go-build-release-windows-amd64 \
+        go-build-release-freebsd-amd64 go-build-release-freebsd-arm64
 
-push:
-	docker push fosrl/newt:latest
+go-build-release: \
+    go-build-release-linux-arm64 \
+    go-build-release-linux-arm32-v7 \
+    go-build-release-linux-arm32-v6 \
+    go-build-release-linux-amd64 \
+    go-build-release-linux-riscv64 \
+    go-build-release-darwin-arm64 \
+    go-build-release-darwin-amd64 \
+    go-build-release-windows-amd64 \
+    go-build-release-freebsd-amd64 \
+    go-build-release-freebsd-arm64
 
-test:
-	docker run fosrl/newt:latest
+go-build-release-linux-arm64:
+	CGO_ENABLED=0 GOOS=linux GOARCH=arm64 go build -ldflags "$(LDFLAGS)" -o bin/newt_linux_arm64
 
-local: 
-	CGO_ENABLED=0 go build -o newt
+go-build-release-linux-arm32-v7:
+	CGO_ENABLED=0 GOOS=linux GOARCH=arm GOARM=7 go build -ldflags "$(LDFLAGS)" -o bin/newt_linux_arm32
 
-go-build-release:
-	CGO_ENABLED=0 GOOS=linux GOARCH=arm64 go build -o bin/newt_linux_arm64
-	CGO_ENABLED=0 GOOS=linux GOARCH=arm GOARM=7 go build -o bin/newt_linux_arm32
-	CGO_ENABLED=0 GOOS=linux GOARCH=arm GOARM=6 go build -o bin/newt_linux_arm32v6
-	CGO_ENABLED=0 GOOS=linux GOARCH=amd64 go build -o bin/newt_linux_amd64
-	CGO_ENABLED=0 GOOS=linux GOARCH=riscv64 go build -o bin/newt_linux_riscv64
-	CGO_ENABLED=0 GOOS=darwin GOARCH=arm64 go build -o bin/newt_darwin_arm64
-	CGO_ENABLED=0 GOOS=darwin GOARCH=amd64 go build -o bin/newt_darwin_amd64
-	CGO_ENABLED=0 GOOS=windows GOARCH=amd64 go build -o bin/newt_windows_amd64.exe
-	CGO_ENABLED=0 GOOS=freebsd GOARCH=amd64 go build -o bin/newt_freebsd_amd64
-	CGO_ENABLED=0 GOOS=freebsd GOARCH=arm64 go build -o bin/newt_freebsd_arm64
+go-build-release-linux-arm32-v6:
+	CGO_ENABLED=0 GOOS=linux GOARCH=arm GOARM=6 go build -ldflags "$(LDFLAGS)" -o bin/newt_linux_arm32v6
 
-clean:
-	rm newt
+go-build-release-linux-amd64:
+	CGO_ENABLED=0 GOOS=linux GOARCH=amd64 go build -ldflags "$(LDFLAGS)" -o bin/newt_linux_amd64
+
+go-build-release-linux-riscv64:
+	CGO_ENABLED=0 GOOS=linux GOARCH=riscv64 go build -ldflags "$(LDFLAGS)" -o bin/newt_linux_riscv64
+
+go-build-release-darwin-arm64:
+	CGO_ENABLED=0 GOOS=darwin GOARCH=arm64 go build -ldflags "$(LDFLAGS)" -o bin/newt_darwin_arm64
+
+go-build-release-darwin-amd64:
+	CGO_ENABLED=0 GOOS=darwin GOARCH=amd64 go build -ldflags "$(LDFLAGS)" -o bin/newt_darwin_amd64
+
+go-build-release-windows-amd64:
+	CGO_ENABLED=0 GOOS=windows GOARCH=amd64 go build -ldflags "$(LDFLAGS)" -o bin/newt_windows_amd64.exe
+
+go-build-release-freebsd-amd64:
+	CGO_ENABLED=0 GOOS=freebsd GOARCH=amd64 go build -ldflags "$(LDFLAGS)" -o bin/newt_freebsd_amd64
+
+go-build-release-freebsd-arm64:
+	CGO_ENABLED=0 GOOS=freebsd GOARCH=arm64 go build -ldflags "$(LDFLAGS)" -o bin/newt_freebsd_arm64

README.md

@@ -9,13 +9,7 @@ Newt is a fully user space [WireGuard](https://www.wireguard.com/) tunnel client
 
 Newt is used with Pangolin and Gerbil as part of the larger system. See documentation below:
 
--   [Full Documentation](https://docs.pangolin.net)
-
-## Preview
-
-<img src="public/screenshots/preview.png" alt="Preview"/>
-
-_Sample output of a Newt connected to Pangolin and hosting various resource target proxies._
+-   [Full Documentation](https://docs.pangolin.net/manage/sites/understanding-sites)
 
 ## Key Functions
 
@@ -31,415 +25,14 @@ When Newt receives WireGuard control messages, it will use the information encod
 
 When Newt receives WireGuard control messages, it will use the information encoded to create a local low level TCP and UDP proxies attached to the virtual tunnel in order to relay traffic to programmed targets.
 
-## CLI Args
-
-### Core Configuration
-
--   `id`: Newt ID generated by Pangolin to identify the client.
--   `secret`: A unique secret (not shared and kept private) used to authenticate the client ID with the websocket in order to receive commands.
--   `endpoint`: The endpoint where both Gerbil and Pangolin reside in order to connect to the websocket.
--   `blueprint-file` (optional): Path to blueprint file to define Pangolin resources and configurations.
--   `no-cloud` (optional): Don't fail over to the cloud when using managed nodes in Pangolin Cloud. Default: false
--   `log-level` (optional): The log level to use (DEBUG, INFO, WARN, ERROR, FATAL). Default: INFO
-
-### Docker Integration
-
--   `docker-socket` (optional): Set the Docker socket to use the container discovery integration
--   `docker-enforce-network-validation` (optional): Validate the container target is on the same network as the newt process. Default: false
-
-### Accpet Client Connection 
-
--   `accept-clients` (optional): Enable WireGuard server mode to accept incoming newt client connections. Default: false
-    -   `generateAndSaveKeyTo` (optional): Path to save generated private key
-    -   `native` (optional): Use native WireGuard interface when accepting clients (requires WireGuard kernel module and Linux, must run as root). Default: false (uses userspace netstack)
-        -   `interface` (optional): Name of the WireGuard interface. Default: newt
-        -   `keep-interface` (optional): Keep the WireGuard interface. Default: false
-
-### Metrics & Observability
-
--   `metrics` (optional): Enable Prometheus /metrics exporter. Default: true
--   `otlp` (optional): Enable OTLP exporters (metrics/traces) to OTEL_EXPORTER_OTLP_ENDPOINT. Default: false
--   `metrics-admin-addr` (optional): Admin/metrics bind address. Default: 127.0.0.1:2112
--   `metrics-async-bytes` (optional): Enable async bytes counting (background flush; lower hot path overhead). Default: false 
--   `region` (optional): Optional region resource attribute for telemetry and metrics.
-
-### Network Configuration
-
--   `mtu` (optional): MTU for the internal WG interface. Default: 1280
--   `dns` (optional): DNS server to use to resolve the endpoint. Default: 9.9.9.9
--   `ping-interval` (optional): Interval for pinging the server. Default: 3s
--   `ping-timeout` (optional): Timeout for each ping. Default: 5s
-
-### Security & TLS
-
--   `enforce-hc-cert` (optional): Enforce certificate validation for health checks. Default: false (accepts any cert)
--   `tls-client-cert` (optional): Client certificate (p12 or pfx) for mTLS or path to client certificate (PEM format). See [mTLS](#mtls)
--   `tls-client-key` (optional): Path to private key for mTLS (PEM format, optional if using PKCS12)
--   `tls-ca-cert` (optional): Path to CA certificate to verify server (PEM format, optional if using PKCS12)
-
-### Monitoring & Health
-
--   `health-file` (optional): Check if connection to WG server (pangolin) is ok. creates a file if ok, removes it if not ok. Can be used with docker healtcheck to restart newt
--   `updown` (optional): A script to be called when targets are added or removed.
-
-## Environment Variables
-
-All CLI arguments can be set using environment variables as an alternative to command line flags. Environment variables are particularly useful when running Newt in containerized environments.
-
-### Core Configuration
-
--   `PANGOLIN_ENDPOINT`: Endpoint of your pangolin server (equivalent to `--endpoint`)
--   `NEWT_ID`: Newt ID generated by Pangolin (equivalent to `--id`)
--   `NEWT_SECRET`: Newt secret for authentication (equivalent to `--secret`)
--   `CONFIG_FILE`: Load the config json from this file instead of in the home folder.
--   `BLUEPRINT_FILE`: Path to blueprint file to define Pangolin resources and configurations. (equivalent to `--blueprint-file`)
--   `NO_CLOUD`: Don't fail over to the cloud when using managed nodes in Pangolin Cloud. Default: false (equivalent to `--no-cloud`)
--   `LOG_LEVEL`: Log level (DEBUG, INFO, WARN, ERROR, FATAL). Default: INFO (equivalent to `--log-level`)
-
-### Docker Integration
-
--   `DOCKER_SOCKET`: Path to Docker socket for container discovery (equivalent to `--docker-socket`)
--   `DOCKER_ENFORCE_NETWORK_VALIDATION`: Validate container targets are on same network. Default: false (equivalent to `--docker-enforce-network-validation`)
-
-### Accept Client Connections
-
--   `ACCEPT_CLIENTS`: Enable WireGuard server mode. Default: false (equivalent to `--accept-clients`)
--   `GENERATE_AND_SAVE_KEY_TO`: Path to save generated private key (equivalent to `--generateAndSaveKeyTo`)
--   `USE_NATIVE_INTERFACE`: Use native WireGuard interface (Linux only). Default: false (equivalent to `--native`)
--   `INTERFACE`: Name of the WireGuard interface. Default: newt (equivalent to `--interface`)
--   `KEEP_INTERFACE`: Keep the WireGuard interface after shutdown. Default: false (equivalent to `--keep-interface`)
-
-### Monitoring & Health
-
--   `HEALTH_FILE`: Path to health file for connection monitoring (equivalent to `--health-file`)
--   `UPDOWN_SCRIPT`: Path to updown script for target add/remove events (equivalent to `--updown`)
-
-### Metrics & Observability
-
--   `NEWT_METRICS_PROMETHEUS_ENABLED`: Enable Prometheus /metrics exporter. Default: true (equivalent to `--metrics`)
--   `NEWT_METRICS_OTLP_ENABLED`: Enable OTLP exporters (metrics/traces) to OTEL_EXPORTER_OTLP_ENDPOINT. Default: false (equivalent to `--otlp`)
--   `NEWT_ADMIN_ADDR`: Admin/metrics bind address. Default: 127.0.0.1:2112 (equivalent to `--metrics-admin-addr`)
--   `NEWT_METRICS_ASYNC_BYTES`: Enable async bytes counting (background flush; lower hot path overhead). Default: false (equivalent to `--metrics-async-bytes`)
--   `NEWT_REGION`: Optional region resource attribute for telemetry and metrics (equivalent to `--region`)
-
-### Network Configuration
-
--   `MTU`: MTU for the internal WG interface. Default: 1280 (equivalent to `--mtu`)
--   `DNS`: DNS server to use to resolve the endpoint. Default: 9.9.9.9 (equivalent to `--dns`)
--   `PING_INTERVAL`: Interval for pinging the server. Default: 3s (equivalent to `--ping-interval`)
--   `PING_TIMEOUT`: Timeout for each ping. Default: 5s (equivalent to `--ping-timeout`)
-
-### Security & TLS
-
--   `ENFORCE_HC_CERT`: Enforce certificate validation for health checks. Default: false (equivalent to `--enforce-hc-cert`)
--   `TLS_CLIENT_CERT`: Path to client certificate for mTLS (equivalent to `--tls-client-cert`)
--   `TLS_CLIENT_KEY`: Path to private key for mTLS (equivalent to `--tls-client-key`)
--   `TLS_CA_CERT`: Path to CA certificate to verify server (equivalent to `--tls-ca-cert`)
--   `SKIP_TLS_VERIFY`: Skip TLS verification for server connections. Default: false
-
-## Loading secrets from files
-
-You can use `CONFIG_FILE` to define a location of a config file to store the credentials between runs. 
-
-```
-$ cat ~/.config/newt-client/config.json
-{
-  "id": "spmzu8rbpzj1qq6",
-  "secret": "f6v61mjutwme2kkydbw3fjo227zl60a2tsf5psw9r25hgae3",
-  "endpoint": "https://app.pangolin.net",
-  "tlsClientCert": ""
-}
-```
-
-This file is also written to when newt first starts up. So you do not need to run every time with --id and secret if you have run it once! 
-
-Default locations: 
-
-- **macOS**: `~/Library/Application Support/newt-client/config.json`
-- **Windows**: `%PROGRAMDATA%\newt\newt-client\config.json`
-- **Linux/Others**: `~/.config/newt-client/config.json`
-
-## Examples
-
-**Note**: When both environment variables and CLI arguments are provided, CLI arguments take precedence.
-
--   Example:
-
-```bash
-newt \
---id 31frd0uzbjvp721 \
---secret h51mmlknrvrwv8s4r1i210azhumt6isgbpyavxodibx1k2d6 \
---endpoint https://example.com
-```
-
-You can also run it with Docker compose. For example, a service in your `docker-compose.yml` might look like this using environment vars (recommended):
-
-```yaml
-services:
-    newt:
-        image: fosrl/newt
-        container_name: newt
-        restart: unless-stopped
-        environment:
-            - PANGOLIN_ENDPOINT=https://example.com
-            - NEWT_ID=2ix2t8xk22ubpfy
-            - NEWT_SECRET=nnisrfsdfc7prqsp9ewo1dvtvci50j5uiqotez00dgap0ii2
-            - HEALTH_FILE=/tmp/healthy
-```
-
-You can also pass the CLI args to the container:
-
-```yaml
-services:
-    newt:
-        image: fosrl/newt
-        container_name: newt
-        restart: unless-stopped
-        command:
-            - --id 31frd0uzbjvp721
-            - --secret h51mmlknrvrwv8s4r1i210azhumt6isgbpyavxodibx1k2d6
-            - --endpoint https://example.com
-            - --health-file /tmp/healthy
-```
-
-## Accept Client Connections
-
-When the `--accept-clients` flag is enabled (or `ACCEPT_CLIENTS=true` environment variable is set), Newt operates as a WireGuard server that can accept incoming client connections from other devices. This enables peer-to-peer connectivity through the Newt instance.
-
-### How It Works
-
-In client acceptance mode, Newt:
-
--   **Creates a WireGuard service** that can accept incoming connections from other WireGuard clients
--   **Starts a connection testing server** (WGTester) that responds to connectivity checks from remote clients
--   **Manages peer configurations** dynamically based on Pangolin's instructions
--   **Enables bidirectional communication** between the Newt instance and connected clients
-
-### Use Cases
-
--   **Site-to-site connectivity**: Connect multiple locations through a central Newt instance
--   **Client access to private networks**: Allow remote clients to access resources behind the Newt instance
--   **Development environments**: Provide developers secure access to internal services
-
-### Client Tunneling Modes
-
-Newt supports two WireGuard tunneling modes:
-
-#### Userspace Mode (Default)
-
-By default, Newt uses a fully userspace WireGuard implementation using [netstack](https://github.com/WireGuard/wireguard-go/blob/master/tun/netstack/examples/http_server.go). This mode:
-
--   **Does not require root privileges**
--   **Works on all supported platforms** (Linux, Windows, macOS)
--   **Does not require WireGuard kernel module** to be installed
--   **Runs entirely in userspace** - no system network interface is created
--   **Is containerization-friendly** - works seamlessly in Docker containers
-
-This is the recommended mode for most deployments, especially containerized environments.
-
-In this mode, TCP and UDP is proxied out of newt from the remote client using TCP/UDP resources in Pangolin.
-
-#### Native Mode (Linux only)
-
-When using the `--native` flag or setting `USE_NATIVE_INTERFACE=true`, Newt uses the native WireGuard kernel module. This mode:
-
--   **Requires root privileges** to create and manage network interfaces
--   **Only works on Linux** with the WireGuard kernel module installed
--   **Creates a real network interface** (e.g., `newt0`) on the system
--   **May offer better performance** for high-throughput scenarios
--   **Requires proper network permissions** and may conflict with existing network configurations
-
-In this mode it functions like a traditional VPN interface - all data arrives on the interface and you must get it to the destination (or access things locally).
-
-#### Native Mode Requirements
-
-To use native mode:
-
-1. Run on a Linux system
-2. Install the WireGuard kernel module
-3. Run Newt as root (`sudo`)
-4. Ensure the system allows creation of network interfaces
-
-Docker Compose example:
-
-```yaml
-services:
-    newt:
-        image: fosrl/newt
-        container_name: newt
-        restart: unless-stopped
-        environment:
-            - PANGOLIN_ENDPOINT=https://example.com
-            - NEWT_ID=2ix2t8xk22ubpfy
-            - NEWT_SECRET=nnisrfsdfc7prqsp9ewo1dvtvci50j5uiqotez00dgap0ii2
-            - ACCEPT_CLIENTS=true
-```
-
-### Technical Details
-
-When client acceptance is enabled:
-
--   **WGTester Server**: Runs on `port + 1` (e.g., if WireGuard uses port 51820, WGTester uses 51821)
--   **Connection Testing**: Responds to UDP packets with magic header `0xDEADBEEF` for connectivity verification
--   **Dynamic Configuration**: Peer configurations are managed remotely through Pangolin
--   **Proxy Integration**: Can work with both userspace (netstack) and native WireGuard modes
-
-**Note**: Client acceptance mode requires coordination with Pangolin for peer management and configuration distribution.
-
-### Docker Socket Integration
-
-Newt can integrate with the Docker socket to provide remote inspection of Docker containers. This allows Pangolin to query and retrieve detailed information about containers running on the Newt client, including metadata, network configuration, port mappings, and more.
-
-**Configuration:**
-
-You can specify the Docker socket path using the `--docker-socket` CLI argument or by setting the `DOCKER_SOCKET` environment variable. If the Docker socket is not available or accessible, Newt will gracefully disable Docker integration and continue normal operation.
-
-Supported values include:
-
--   Local UNIX socket (default):
-    >You must mount the socket file into the container using a volume, so Newt can access it.
-
-    `unix:///var/run/docker.sock`
-
--   TCP socket (e.g., via Docker Socket Proxy):
-
-    `tcp://localhost:2375`
-
--   HTTP/HTTPS endpoints (e.g., remote Docker APIs):
-
-    `http://your-host:2375`
-
--   SSH connections (experimental, requires SSH setup):
-
-    `ssh://user@host`
-
-
-```yaml
-services:
-    newt:
-        image: fosrl/newt
-        container_name: newt
-        restart: unless-stopped
-        volumes:
-            - /var/run/docker.sock:/var/run/docker.sock:ro
-        environment:
-            - PANGOLIN_ENDPOINT=https://example.com
-            - NEWT_ID=2ix2t8xk22ubpfy
-            - NEWT_SECRET=nnisrfsdfc7prqsp9ewo1dvtvci50j5uiqotez00dgap0ii2
-            - DOCKER_SOCKET=unix:///var/run/docker.sock
-```
->If you previously used just a path like `/var/run/docker.sock`, it still works — Newt assumes it is a UNIX socket by default.
-
-#### Hostnames vs IPs
-
-When the Docker Socket Integration is used, depending on the network which Newt is run with, either the hostname (generally considered the container name) or the IP address of the container will be sent to Pangolin. Here are some of the scenarios where IPs or hostname of the container will be utilised:
-
--   **Running in Network Mode 'host'**: IP addresses will be used
--   **Running in Network Mode 'bridge'**: IP addresses will be used
--   **Running in docker-compose without a network specification**: Docker compose creates a network for the compose by default, hostnames will be used
--   **Running on docker-compose with defined network**: Hostnames will be used
-
-### Docker Enforce Network Validation
-
-When run as a Docker container, Newt can validate that the target being provided is on the same network as the Newt container and only return containers directly accessible by Newt. Validation will be carried out against either the hostname/IP Address and the Port number to ensure the running container is exposing the ports to Newt.
-
-It is important to note that if the Newt container is run with a network mode of `host` that this feature will not work. Running in `host` mode causes the container to share its resources with the host machine, therefore making it so the specific host container information for Newt cannot be retrieved to be able to carry out network validation.
-
-**Configuration:**
-
-Validation is `false` by default. It can be enabled via setting the `--docker-enforce-network-validation` CLI argument or by setting the `DOCKER_ENFORCE_NETWORK_VALIDATION` environment variable.
-
-If validation is enforced and the Docker socket is available, Newt will **not** add the target as it cannot be verified. A warning will be presented in the Newt logs.
-
-### Updown
-
-You can pass in a updown script for Newt to call when it is adding or removing a target:
-
-`--updown "python3 test.py"`
-
-It will get called with args when a target is added:
-`python3 test.py add tcp localhost:8556`
-`python3 test.py remove tcp localhost:8556`
-
-Returning a string from the script in the format of a target (`ip:dst` so `10.0.0.1:8080`) it will override the target and use this value instead to proxy.
-
-You can look at updown.py as a reference script to get started!
-
-### mTLS
-
-Newt supports mutual TLS (mTLS) authentication if the server is configured to request a client certificate. You can use either a PKCS12 (.p12/.pfx) file or split PEM files for the client cert, private key, and CA.
-
-#### Option 1: PKCS12 (Legacy)
-
-> This is the original method and still supported.
-
-* File must contain:
-
-  * Client private key
-  * Public certificate
-  * CA certificate
-* Encrypted `.p12` files are **not supported**
-
-Example:
-
-```bash
-newt \
---id 31frd0uzbjvp721 \
---secret h51mmlknrvrwv8s4r1i210azhumt6isgbpyavxodibx1k2d6 \
---endpoint https://example.com \
---tls-client-cert ./client.p12
-```
-
-#### Option 2: Split PEM Files (Preferred)
-
-You can now provide separate files for:
-
-* `--tls-client-cert`: client certificate (`.crt` or `.pem`)
-* `--tls-client-key`: client private key (`.key` or `.pem`)
-* `--tls-ca-cert`: CA cert to verify the server
-
-Example:
-
-```bash
-newt \
---id 31frd0uzbjvp721 \
---secret h51mmlknrvrwv8s4r1i210azhumt6isgbpyavxodibx1k2d6 \
---endpoint https://example.com \
---tls-client-cert ./client.crt \
---tls-client-key ./client.key \
---tls-ca-cert ./ca.crt
-```
-
-
-```yaml
-services:
-    newt:
-        image: fosrl/newt
-        container_name: newt
-        restart: unless-stopped
-        environment:
-            - PANGOLIN_ENDPOINT=https://example.com
-            - NEWT_ID=2ix2t8xk22ubpfy
-            - NEWT_SECRET=nnisrfsdfc7prqsp9ewo1dvtvci50j5uiqotez00dgap0ii2
-            - TLS_CLIENT_CERT=./client.p12
-```
-
 ## Build
 
-### Container
-
-Ensure Docker is installed.
-
-```bash
-make
-```
-
 ### Binary
 
-Make sure to have Go 1.23.1 installed.
+Make sure to have Go 1.25 installed.
 
 ```bash
-make local
+make
 ```
 
 ### Nix Flake

authdaemon.go

@@ -0,0 +1,150 @@
+package main
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"os"
+	"runtime"
+
+	"github.com/fosrl/newt/authdaemon"
+	"github.com/fosrl/newt/logger"
+)
+
+const (
+	defaultPrincipalsPath = "/var/run/auth-daemon/principals"
+	defaultCACertPath     = "/etc/ssh/ca.pem"
+)
+
+var (
+	errPresharedKeyRequired = errors.New("auth-daemon-key is required when --auth-daemon is enabled")
+	errRootRequired         = errors.New("auth-daemon must be run as root (use sudo)")
+	authDaemonServer        *authdaemon.Server // Global auth daemon server instance
+)
+
+// startAuthDaemon initializes and starts the auth daemon in the background.
+// It validates requirements (Linux, root, preshared key) and starts the server
+// in a goroutine so it runs alongside normal newt operation.
+func startAuthDaemon(ctx context.Context) error {
+	// Validation
+	if runtime.GOOS != "linux" {
+		return fmt.Errorf("auth-daemon is only supported on Linux, not %s", runtime.GOOS)
+	}
+	if os.Geteuid() != 0 {
+		return errRootRequired
+	}
+
+	// Use defaults if not set
+	principalsFile := authDaemonPrincipalsFile
+	if principalsFile == "" {
+		principalsFile = defaultPrincipalsPath
+	}
+	caCertPath := authDaemonCACertPath
+	if caCertPath == "" {
+		caCertPath = defaultCACertPath
+	}
+
+	// Create auth daemon server
+	cfg := authdaemon.Config{
+		DisableHTTPS:           true,                   // We run without HTTP server in newt
+		PresharedKey:           "this-key-is-not-used", // Not used in embedded mode, but set to non-empty to satisfy validation
+		PrincipalsFilePath:     principalsFile,
+		CACertPath:             caCertPath,
+		Force:                  true,
+		GenerateRandomPassword: authDaemonGenerateRandomPassword,
+	}
+
+	srv, err := authdaemon.NewServer(cfg)
+	if err != nil {
+		return fmt.Errorf("create auth daemon server: %w", err)
+	}
+
+	authDaemonServer = srv
+
+	// Start the auth daemon in a goroutine so it runs alongside newt
+	go func() {
+		logger.Info("Auth daemon starting (native mode, no HTTP server)")
+		if err := srv.Run(ctx); err != nil {
+			logger.Error("Auth daemon error: %v", err)
+		}
+		logger.Info("Auth daemon stopped")
+	}()
+
+	return nil
+}
+
+// runPrincipalsCmd executes the principals subcommand logic
+func runPrincipalsCmd(args []string) {
+	opts := struct {
+		PrincipalsFile string
+		Username       string
+	}{
+		PrincipalsFile: defaultPrincipalsPath,
+	}
+
+	// Parse flags manually
+	for i := 0; i < len(args); i++ {
+		switch args[i] {
+		case "--principals-file":
+			if i+1 >= len(args) {
+				fmt.Fprintf(os.Stderr, "Error: --principals-file requires a value\n")
+				os.Exit(1)
+			}
+			opts.PrincipalsFile = args[i+1]
+			i++
+		case "--username":
+			if i+1 >= len(args) {
+				fmt.Fprintf(os.Stderr, "Error: --username requires a value\n")
+				os.Exit(1)
+			}
+			opts.Username = args[i+1]
+			i++
+		case "--help", "-h":
+			printPrincipalsHelp()
+			os.Exit(0)
+		default:
+			fmt.Fprintf(os.Stderr, "Error: unknown flag: %s\n", args[i])
+			printPrincipalsHelp()
+			os.Exit(1)
+		}
+	}
+
+	// Validation
+	if opts.Username == "" {
+		fmt.Fprintf(os.Stderr, "Error: username is required\n")
+		printPrincipalsHelp()
+		os.Exit(1)
+	}
+
+	// Get principals
+	list, err := authdaemon.GetPrincipals(opts.PrincipalsFile, opts.Username)
+	if err != nil {
+		fmt.Fprintf(os.Stderr, "Error: %v\n", err)
+		os.Exit(1)
+	}
+	if len(list) == 0 {
+		fmt.Println("")
+		return
+	}
+	for _, principal := range list {
+		fmt.Println(principal)
+	}
+}
+
+func printPrincipalsHelp() {
+	fmt.Fprintf(os.Stderr, `Usage: newt principals [flags]
+
+Output principals for a username (for AuthorizedPrincipalsCommand in sshd_config).
+Read the principals file and print principals that match the given username, one per line.
+Configure in sshd_config with AuthorizedPrincipalsCommand and %%u for the username.
+
+Flags:
+  --principals-file string   Path to the principals file (default "%s")
+  --username string          Username to look up (required)
+  --help, -h                 Show this help message
+
+Example:
+  newt principals --username alice
+
+`, defaultPrincipalsPath)
+}

authdaemon/connection.go

@@ -0,0 +1,27 @@
+package authdaemon
+
+import (
+	"github.com/fosrl/newt/logger"
+)
+
+// ProcessConnection runs the same logic as POST /connection: CA cert, user create/reconcile, principals.
+// Use this when DisableHTTPS is true (e.g. embedded in Newt) instead of calling the API.
+func (s *Server) ProcessConnection(req ConnectionRequest) {
+	logger.Info("connection: niceId=%q username=%q metadata.sudoMode=%q metadata.sudoCommands=%v metadata.homedir=%v metadata.groups=%v",
+		req.NiceId, req.Username, req.Metadata.SudoMode, req.Metadata.SudoCommands, req.Metadata.Homedir, req.Metadata.Groups)
+
+	cfg := &s.cfg
+	if cfg.CACertPath != "" {
+		if err := writeCACertIfNotExists(cfg.CACertPath, req.CaCert, cfg.Force); err != nil {
+			logger.Warn("auth-daemon: write CA cert: %v", err)
+		}
+	}
+	if err := ensureUser(req.Username, req.Metadata, s.cfg.GenerateRandomPassword); err != nil {
+		logger.Warn("auth-daemon: ensure user: %v", err)
+	}
+	if cfg.PrincipalsFilePath != "" {
+		if err := writePrincipals(cfg.PrincipalsFilePath, req.Username, req.NiceId); err != nil {
+			logger.Warn("auth-daemon: write principals: %v", err)
+		}
+	}
+}

authdaemon/host_linux.go

@@ -0,0 +1,396 @@
+//go:build linux
+
+package authdaemon
+
+import (
+	"bufio"
+	"crypto/rand"
+	"encoding/hex"
+	"encoding/json"
+	"fmt"
+	"os"
+	"os/exec"
+	"os/user"
+	"path/filepath"
+	"strconv"
+	"strings"
+
+	"github.com/fosrl/newt/logger"
+)
+
+// writeCACertIfNotExists writes contents to path. If the file already exists: when force is false, skip; when force is true, overwrite only if content differs.
+func writeCACertIfNotExists(path, contents string, force bool) error {
+	contents = strings.TrimSpace(contents)
+	if contents != "" && !strings.HasSuffix(contents, "\n") {
+		contents += "\n"
+	}
+	existing, err := os.ReadFile(path)
+	if err == nil {
+		existingStr := strings.TrimSpace(string(existing))
+		if existingStr != "" && !strings.HasSuffix(existingStr, "\n") {
+			existingStr += "\n"
+		}
+		if existingStr == contents {
+			logger.Debug("auth-daemon: CA cert unchanged at %s, skipping write", path)
+			return nil
+		}
+		if !force {
+			logger.Debug("auth-daemon: CA cert already exists at %s, skipping write (Force disabled)", path)
+			return nil
+		}
+	} else if !os.IsNotExist(err) {
+		return fmt.Errorf("read %s: %w", path, err)
+	}
+	dir := filepath.Dir(path)
+	if err := os.MkdirAll(dir, 0755); err != nil {
+		return fmt.Errorf("mkdir %s: %w", dir, err)
+	}
+	if err := os.WriteFile(path, []byte(contents), 0644); err != nil {
+		return fmt.Errorf("write CA cert: %w", err)
+	}
+	logger.Info("auth-daemon: wrote CA cert to %s", path)
+	return nil
+}
+
+// writePrincipals updates the principals file at path: JSON object keyed by username, value is array of principals. Adds username and niceId to that user's list (deduped).
+func writePrincipals(path, username, niceId string) error {
+	if path == "" {
+		return nil
+	}
+	username = strings.TrimSpace(username)
+	niceId = strings.TrimSpace(niceId)
+	if username == "" {
+		return nil
+	}
+	data := make(map[string][]string)
+	if raw, err := os.ReadFile(path); err == nil {
+		_ = json.Unmarshal(raw, &data)
+	}
+	list := data[username]
+	seen := make(map[string]struct{}, len(list)+2)
+	for _, p := range list {
+		seen[p] = struct{}{}
+	}
+	for _, p := range []string{username, niceId} {
+		if p == "" {
+			continue
+		}
+		if _, ok := seen[p]; !ok {
+			seen[p] = struct{}{}
+			list = append(list, p)
+		}
+	}
+	data[username] = list
+	body, err := json.Marshal(data)
+	if err != nil {
+		return fmt.Errorf("marshal principals: %w", err)
+	}
+	dir := filepath.Dir(path)
+	if err := os.MkdirAll(dir, 0755); err != nil {
+		return fmt.Errorf("mkdir %s: %w", dir, err)
+	}
+	if err := os.WriteFile(path, body, 0644); err != nil {
+		return fmt.Errorf("write principals: %w", err)
+	}
+	logger.Debug("auth-daemon: wrote principals to %s", path)
+	return nil
+}
+
+// sudoGroup returns the name of the sudo group (wheel or sudo) that exists on the system. Prefers wheel.
+func sudoGroup() string {
+	f, err := os.Open("/etc/group")
+	if err != nil {
+		return "sudo"
+	}
+	defer f.Close()
+	sc := bufio.NewScanner(f)
+	hasWheel := false
+	hasSudo := false
+	for sc.Scan() {
+		line := sc.Text()
+		if strings.HasPrefix(line, "wheel:") {
+			hasWheel = true
+		}
+		if strings.HasPrefix(line, "sudo:") {
+			hasSudo = true
+		}
+	}
+	if hasWheel {
+		return "wheel"
+	}
+	if hasSudo {
+		return "sudo"
+	}
+	return "sudo"
+}
+
+// setRandomPassword generates a random password and sets it for username via chpasswd.
+// Used when GenerateRandomPassword is true so SSH with PermitEmptyPasswords no can accept the user.
+func setRandomPassword(username string) error {
+	b := make([]byte, 16)
+	if _, err := rand.Read(b); err != nil {
+		return fmt.Errorf("generate password: %w", err)
+	}
+	password := hex.EncodeToString(b)
+	cmd := exec.Command("chpasswd")
+	cmd.Stdin = strings.NewReader(username + ":" + password)
+	if out, err := cmd.CombinedOutput(); err != nil {
+		return fmt.Errorf("chpasswd: %w (output: %s)", err, string(out))
+	}
+	return nil
+}
+
+const skelDir = "/etc/skel"
+
+// copySkelInto copies files from srcDir (e.g. /etc/skel) into dstDir (e.g. user's home).
+// Only creates files that don't already exist. All created paths are chowned to uid:gid.
+func copySkelInto(srcDir, dstDir string, uid, gid int) {
+	entries, err := os.ReadDir(srcDir)
+	if err != nil {
+		if !os.IsNotExist(err) {
+			logger.Warn("auth-daemon: read %s: %v", srcDir, err)
+		}
+		return
+	}
+	for _, e := range entries {
+		name := e.Name()
+		src := filepath.Join(srcDir, name)
+		dst := filepath.Join(dstDir, name)
+		if e.IsDir() {
+			if st, err := os.Stat(dst); err == nil && st.IsDir() {
+				copySkelInto(src, dst, uid, gid)
+				continue
+			}
+			if err := os.MkdirAll(dst, 0755); err != nil {
+				logger.Warn("auth-daemon: mkdir %s: %v", dst, err)
+				continue
+			}
+			if err := os.Chown(dst, uid, gid); err != nil {
+				logger.Warn("auth-daemon: chown %s: %v", dst, err)
+			}
+			copySkelInto(src, dst, uid, gid)
+			continue
+		}
+		if _, err := os.Stat(dst); err == nil {
+			continue
+		}
+		data, err := os.ReadFile(src)
+		if err != nil {
+			logger.Warn("auth-daemon: read %s: %v", src, err)
+			continue
+		}
+		if err := os.WriteFile(dst, data, 0644); err != nil {
+			logger.Warn("auth-daemon: write %s: %v", dst, err)
+			continue
+		}
+		if err := os.Chown(dst, uid, gid); err != nil {
+			logger.Warn("auth-daemon: chown %s: %v", dst, err)
+		}
+	}
+}
+
+// ensureUser creates the system user if missing, or reconciles sudo and homedir to match meta.
+func ensureUser(username string, meta ConnectionMetadata, generateRandomPassword bool) error {
+	if username == "" {
+		return nil
+	}
+	u, err := user.Lookup(username)
+	if err != nil {
+		if _, ok := err.(user.UnknownUserError); !ok {
+			return fmt.Errorf("lookup user %s: %w", username, err)
+		}
+		return createUser(username, meta, generateRandomPassword)
+	}
+	return reconcileUser(u, meta)
+}
+
+// desiredGroups returns the exact list of supplementary groups the user should have:
+// meta.Groups plus the sudo group when meta.SudoMode is "full" (deduped).
+func desiredGroups(meta ConnectionMetadata) []string {
+	seen := make(map[string]struct{})
+	var out []string
+	for _, g := range meta.Groups {
+		g = strings.TrimSpace(g)
+		if g == "" {
+			continue
+		}
+		if _, ok := seen[g]; ok {
+			continue
+		}
+		seen[g] = struct{}{}
+		out = append(out, g)
+	}
+	if meta.SudoMode == "full" {
+		sg := sudoGroup()
+		if _, ok := seen[sg]; !ok {
+			out = append(out, sg)
+		}
+	}
+	return out
+}
+
+// setUserGroups sets the user's supplementary groups to exactly groups (local mirrors metadata).
+// When groups is empty, clears all supplementary groups (usermod -G "").
+func setUserGroups(username string, groups []string) {
+	list := strings.Join(groups, ",")
+	cmd := exec.Command("usermod", "-G", list, username)
+	if out, err := cmd.CombinedOutput(); err != nil {
+		logger.Warn("auth-daemon: usermod -G %s: %v (output: %s)", list, err, string(out))
+	} else {
+		logger.Info("auth-daemon: set %s supplementary groups to %s", username, list)
+	}
+}
+
+func createUser(username string, meta ConnectionMetadata, generateRandomPassword bool) error {
+	args := []string{"-s", "/bin/bash"}
+	if meta.Homedir {
+		args = append(args, "-m")
+	} else {
+		args = append(args, "-M")
+	}
+	args = append(args, username)
+	cmd := exec.Command("useradd", args...)
+	if out, err := cmd.CombinedOutput(); err != nil {
+		return fmt.Errorf("useradd %s: %w (output: %s)", username, err, string(out))
+	}
+	logger.Info("auth-daemon: created user %s (homedir=%v)", username, meta.Homedir)
+	if generateRandomPassword {
+		if err := setRandomPassword(username); err != nil {
+			logger.Warn("auth-daemon: set random password for %s: %v", username, err)
+		} else {
+			logger.Info("auth-daemon: set random password for %s (PermitEmptyPasswords no)", username)
+		}
+	}
+	if meta.Homedir {
+		if u, err := user.Lookup(username); err == nil && u.HomeDir != "" {
+			uid, gid := mustAtoi(u.Uid), mustAtoi(u.Gid)
+			copySkelInto(skelDir, u.HomeDir, uid, gid)
+		}
+	}
+	setUserGroups(username, desiredGroups(meta))
+	switch meta.SudoMode {
+	case "full":
+		if err := configurePasswordlessSudo(username); err != nil {
+			logger.Warn("auth-daemon: configure passwordless sudo for %s: %v", username, err)
+		}
+	case "commands":
+		if len(meta.SudoCommands) > 0 {
+			if err := configureSudoCommands(username, meta.SudoCommands); err != nil {
+				logger.Warn("auth-daemon: configure sudo commands for %s: %v", username, err)
+			}
+		}
+	default:
+		removeSudoers(username)
+	}
+	return nil
+}
+
+const sudoersFilePrefix = "90-pangolin-"
+
+func sudoersPath(username string) string {
+	return filepath.Join("/etc/sudoers.d", sudoersFilePrefix+username)
+}
+
+// writeSudoersFile writes content to the user's sudoers.d file and validates with visudo.
+func writeSudoersFile(username, content string) error {
+	sudoersFile := sudoersPath(username)
+	tmpFile := sudoersFile + ".tmp"
+	if err := os.WriteFile(tmpFile, []byte(content), 0440); err != nil {
+		return fmt.Errorf("write temp sudoers file: %w", err)
+	}
+	cmd := exec.Command("visudo", "-c", "-f", tmpFile)
+	if out, err := cmd.CombinedOutput(); err != nil {
+		os.Remove(tmpFile)
+		return fmt.Errorf("visudo validation failed: %w (output: %s)", err, string(out))
+	}
+	if err := os.Rename(tmpFile, sudoersFile); err != nil {
+		os.Remove(tmpFile)
+		return fmt.Errorf("move sudoers file: %w", err)
+	}
+	return nil
+}
+
+// configurePasswordlessSudo creates a sudoers.d file to allow passwordless sudo for the user.
+func configurePasswordlessSudo(username string) error {
+	content := fmt.Sprintf("# Created by Pangolin auth-daemon\n%s ALL=(ALL) NOPASSWD:ALL\n", username)
+	if err := writeSudoersFile(username, content); err != nil {
+		return err
+	}
+	logger.Info("auth-daemon: configured passwordless sudo for %s", username)
+	return nil
+}
+
+// configureSudoCommands creates a sudoers.d file allowing only the listed commands (NOPASSWD).
+// Each command should be a full path (e.g. /usr/bin/systemctl).
+func configureSudoCommands(username string, commands []string) error {
+	var b strings.Builder
+	b.WriteString("# Created by Pangolin auth-daemon (restricted commands)\n")
+	n := 0
+	for _, c := range commands {
+		c = strings.TrimSpace(c)
+		if c == "" {
+			continue
+		}
+		fmt.Fprintf(&b, "%s ALL=(ALL) NOPASSWD: %s\n", username, c)
+		n++
+	}
+	if n == 0 {
+		return fmt.Errorf("no valid sudo commands")
+	}
+	if err := writeSudoersFile(username, b.String()); err != nil {
+		return err
+	}
+	logger.Info("auth-daemon: configured restricted sudo for %s (%d commands)", username, len(commands))
+	return nil
+}
+
+// removeSudoers removes the sudoers.d file for the user.
+func removeSudoers(username string) {
+	sudoersFile := sudoersPath(username)
+	if err := os.Remove(sudoersFile); err != nil && !os.IsNotExist(err) {
+		logger.Warn("auth-daemon: remove sudoers for %s: %v", username, err)
+	} else if err == nil {
+		logger.Info("auth-daemon: removed sudoers for %s", username)
+	}
+}
+
+func mustAtoi(s string) int {
+	n, _ := strconv.Atoi(s)
+	return n
+}
+
+func reconcileUser(u *user.User, meta ConnectionMetadata) error {
+	setUserGroups(u.Username, desiredGroups(meta))
+	switch meta.SudoMode {
+	case "full":
+		if err := configurePasswordlessSudo(u.Username); err != nil {
+			logger.Warn("auth-daemon: configure passwordless sudo for %s: %v", u.Username, err)
+		}
+	case "commands":
+		if len(meta.SudoCommands) > 0 {
+			if err := configureSudoCommands(u.Username, meta.SudoCommands); err != nil {
+				logger.Warn("auth-daemon: configure sudo commands for %s: %v", u.Username, err)
+			}
+		} else {
+			removeSudoers(u.Username)
+		}
+	default:
+		removeSudoers(u.Username)
+	}
+	if meta.Homedir && u.HomeDir != "" {
+		uid, gid := mustAtoi(u.Uid), mustAtoi(u.Gid)
+		if st, err := os.Stat(u.HomeDir); err != nil || !st.IsDir() {
+			if err := os.MkdirAll(u.HomeDir, 0755); err != nil {
+				logger.Warn("auth-daemon: mkdir %s: %v", u.HomeDir, err)
+			} else {
+				_ = os.Chown(u.HomeDir, uid, gid)
+				copySkelInto(skelDir, u.HomeDir, uid, gid)
+				logger.Info("auth-daemon: created home %s for %s", u.HomeDir, u.Username)
+			}
+		} else {
+			// Ensure .bashrc etc. exist (e.g. home existed but was empty or skel was minimal)
+			copySkelInto(skelDir, u.HomeDir, uid, gid)
+		}
+	}
+	return nil
+}

authdaemon/host_stub.go

@@ -0,0 +1,22 @@
+//go:build !linux
+
+package authdaemon
+
+import "fmt"
+
+var errLinuxOnly = fmt.Errorf("auth-daemon PAM agent is only supported on Linux")
+
+// writeCACertIfNotExists returns an error on non-Linux.
+func writeCACertIfNotExists(path, contents string, force bool) error {
+	return errLinuxOnly
+}
+
+// ensureUser returns an error on non-Linux.
+func ensureUser(username string, meta ConnectionMetadata, generateRandomPassword bool) error {
+	return errLinuxOnly
+}
+
+// writePrincipals returns an error on non-Linux.
+func writePrincipals(path, username, niceId string) error {
+	return errLinuxOnly
+}

authdaemon/principals.go

@@ -0,0 +1,28 @@
+package authdaemon
+
+import (
+	"encoding/json"
+	"fmt"
+	"os"
+)
+
+// GetPrincipals reads the principals data file at path, looks up the given user, and returns that user's principals as a string slice.
+// The file format is JSON: object with username keys and array-of-principals values, e.g. {"alice":["alice","usr-123"],"bob":["bob","usr-456"]}.
+// If the user is not found or the file is missing, returns nil and nil.
+func GetPrincipals(path, user string) ([]string, error) {
+	if path == "" {
+		return nil, fmt.Errorf("principals file path is required")
+	}
+	data, err := os.ReadFile(path)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return nil, nil
+		}
+		return nil, fmt.Errorf("read principals file: %w", err)
+	}
+	var m map[string][]string
+	if err := json.Unmarshal(data, &m); err != nil {
+		return nil, fmt.Errorf("parse principals file: %w", err)
+	}
+	return m[user], nil
+}

authdaemon/routes.go

@@ -0,0 +1,58 @@
+package authdaemon
+
+import (
+	"encoding/json"
+	"net/http"
+)
+
+// registerRoutes registers all API routes. Add new endpoints here.
+func (s *Server) registerRoutes() {
+	s.mux.HandleFunc("/health", s.handleHealth)
+	s.mux.HandleFunc("/connection", s.handleConnection)
+}
+
+// ConnectionMetadata is the metadata object in POST /connection.
+type ConnectionMetadata struct {
+	SudoMode     string   `json:"sudoMode"`     // "none" | "full" | "commands"
+	SudoCommands []string `json:"sudoCommands"`  // used when sudoMode is "commands"
+	Homedir      bool     `json:"homedir"`
+	Groups       []string `json:"groups"`        // system groups to add the user to
+}
+
+// ConnectionRequest is the JSON body for POST /connection.
+type ConnectionRequest struct {
+	CaCert   string             `json:"caCert"`
+	NiceId   string             `json:"niceId"`
+	Username string             `json:"username"`
+	Metadata ConnectionMetadata `json:"metadata"`
+}
+
+// healthResponse is the JSON body for GET /health.
+type healthResponse struct {
+	Status string `json:"status"`
+}
+
+// handleHealth responds with 200 and {"status":"ok"}.
+func (s *Server) handleHealth(w http.ResponseWriter, r *http.Request) {
+	if r.Method != http.MethodGet {
+		http.Error(w, "Method Not Allowed", http.StatusMethodNotAllowed)
+		return
+	}
+	w.Header().Set("Content-Type", "application/json")
+	_ = json.NewEncoder(w).Encode(healthResponse{Status: "ok"})
+}
+
+// handleConnection accepts POST with connection payload and delegates to ProcessConnection.
+func (s *Server) handleConnection(w http.ResponseWriter, r *http.Request) {
+	if r.Method != http.MethodPost {
+		http.Error(w, "Method Not Allowed", http.StatusMethodNotAllowed)
+		return
+	}
+	var req ConnectionRequest
+	if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
+		http.Error(w, "Bad Request", http.StatusBadRequest)
+		return
+	}
+	s.ProcessConnection(req)
+	w.WriteHeader(http.StatusOK)
+}

authdaemon/server.go

@@ -0,0 +1,180 @@
+package authdaemon
+
+import (
+	"context"
+	"crypto/ecdsa"
+	"crypto/elliptic"
+	"crypto/rand"
+	"crypto/subtle"
+	"crypto/tls"
+	"crypto/x509"
+	"crypto/x509/pkix"
+	"encoding/pem"
+	"fmt"
+	"math/big"
+	"net/http"
+	"os"
+	"runtime"
+	"strings"
+	"time"
+
+	"github.com/fosrl/newt/logger"
+)
+
+type Config struct {
+	// DisableHTTPS: when true, Run() does not start the HTTPS server (for embedded use inside Newt). Call ProcessConnection directly for connection events.
+	DisableHTTPS       bool
+	Port               int    // Required when DisableHTTPS is false. Listen port for the HTTPS server. No default.
+	PresharedKey       string // Required when DisableHTTPS is false. HTTP auth (Authorization: Bearer <key> or X-Preshared-Key: <key>). No default.
+	CACertPath         string // Required. Where to write the CA cert (e.g. /etc/ssh/ca.pem). No default.
+	Force                   bool   // If true, overwrite existing CA cert (and other items) when content differs. Default false.
+	PrincipalsFilePath      string // Required. Path to the principals data file (JSON: username -> array of principals). No default.
+	GenerateRandomPassword  bool   // If true, set a random password on users when they are provisioned (for SSH PermitEmptyPasswords no).
+}
+
+type Server struct {
+	cfg          Config
+	addr         string
+	presharedKey string
+	mux          *http.ServeMux
+	tlsCert      tls.Certificate
+}
+
+// generateTLSCert creates a self-signed certificate and key in memory (no disk).
+func generateTLSCert() (tls.Certificate, error) {
+	key, err := ecdsa.GenerateKey(elliptic.P256(), rand.Reader)
+	if err != nil {
+		return tls.Certificate{}, fmt.Errorf("generate key: %w", err)
+	}
+	serial, err := rand.Int(rand.Reader, new(big.Int).Lsh(big.NewInt(1), 128))
+	if err != nil {
+		return tls.Certificate{}, fmt.Errorf("serial: %w", err)
+	}
+	tmpl := &x509.Certificate{
+		SerialNumber: serial,
+		Subject: pkix.Name{
+			CommonName: "localhost",
+		},
+		NotBefore:             time.Now(),
+		NotAfter:              time.Now().Add(365 * 24 * time.Hour),
+		KeyUsage:              x509.KeyUsageDigitalSignature,
+		ExtKeyUsage:           []x509.ExtKeyUsage{x509.ExtKeyUsageServerAuth},
+		BasicConstraintsValid: true,
+		DNSNames:              []string{"localhost", "127.0.0.1"},
+	}
+	certDER, err := x509.CreateCertificate(rand.Reader, tmpl, tmpl, key.Public(), key)
+	if err != nil {
+		return tls.Certificate{}, fmt.Errorf("create certificate: %w", err)
+	}
+	keyDER, err := x509.MarshalECPrivateKey(key)
+	if err != nil {
+		return tls.Certificate{}, fmt.Errorf("marshal key: %w", err)
+	}
+	certPEM := pem.EncodeToMemory(&pem.Block{Type: "CERTIFICATE", Bytes: certDER})
+	keyPEM := pem.EncodeToMemory(&pem.Block{Type: "EC PRIVATE KEY", Bytes: keyDER})
+	cert, err := tls.X509KeyPair(certPEM, keyPEM)
+	if err != nil {
+		return tls.Certificate{}, fmt.Errorf("x509 key pair: %w", err)
+	}
+	return cert, nil
+}
+
+// authMiddleware wraps next and requires a valid preshared key on every request.
+// Accepts Authorization: Bearer <key> or X-Preshared-Key: <key>.
+func (s *Server) authMiddleware(next http.Handler) http.Handler {
+	return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		key := ""
+		if v := r.Header.Get("Authorization"); strings.HasPrefix(v, "Bearer ") {
+			key = strings.TrimSpace(strings.TrimPrefix(v, "Bearer "))
+		}
+		if key == "" {
+			key = strings.TrimSpace(r.Header.Get("X-Preshared-Key"))
+		}
+		if key == "" || subtle.ConstantTimeCompare([]byte(key), []byte(s.presharedKey)) != 1 {
+			http.Error(w, "Unauthorized", http.StatusUnauthorized)
+			return
+		}
+		next.ServeHTTP(w, r)
+	})
+}
+
+// NewServer builds a new auth-daemon server from cfg. Port, PresharedKey, CACertPath, and PrincipalsFilePath are required (no defaults).
+func NewServer(cfg Config) (*Server, error) {
+	if runtime.GOOS != "linux" {
+		return nil, fmt.Errorf("auth-daemon is only supported on Linux, not %s", runtime.GOOS)
+	}
+	if !cfg.DisableHTTPS {
+		if cfg.Port <= 0 {
+			return nil, fmt.Errorf("port is required and must be positive")
+		}
+		if cfg.PresharedKey == "" {
+			return nil, fmt.Errorf("preshared key is required")
+		}
+	}
+	if cfg.CACertPath == "" {
+		return nil, fmt.Errorf("CACertPath is required")
+	}
+	if cfg.PrincipalsFilePath == "" {
+		return nil, fmt.Errorf("PrincipalsFilePath is required")
+	}
+	s := &Server{cfg: cfg}
+	if !cfg.DisableHTTPS {
+		cert, err := generateTLSCert()
+		if err != nil {
+			return nil, err
+		}
+		s.addr = fmt.Sprintf(":%d", cfg.Port)
+		s.presharedKey = cfg.PresharedKey
+		s.mux = http.NewServeMux()
+		s.tlsCert = cert
+		s.registerRoutes()
+	}
+	return s, nil
+}
+
+// Run starts the HTTPS server (unless DisableHTTPS) and blocks until ctx is cancelled or the server errors.
+// When DisableHTTPS is true, Run() blocks on ctx only and does not listen; use ProcessConnection for connection events.
+func (s *Server) Run(ctx context.Context) error {
+	if s.cfg.DisableHTTPS {
+		logger.Info("auth-daemon running (HTTPS disabled)")
+		<-ctx.Done()
+		s.cleanupPrincipalsFile()
+		return nil
+	}
+	tcfg := &tls.Config{
+		Certificates: []tls.Certificate{s.tlsCert},
+		MinVersion:   tls.VersionTLS12,
+	}
+	handler := s.authMiddleware(s.mux)
+	srv := &http.Server{
+		Addr:              s.addr,
+		Handler:           handler,
+		TLSConfig:         tcfg,
+		ReadTimeout:       10 * time.Second,
+		WriteTimeout:      10 * time.Second,
+		ReadHeaderTimeout: 5 * time.Second,
+		IdleTimeout:       60 * time.Second,
+	}
+	go func() {
+		<-ctx.Done()
+		shutdownCtx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+		defer cancel()
+		if err := srv.Shutdown(shutdownCtx); err != nil {
+			logger.Warn("auth-daemon shutdown: %v", err)
+		}
+	}()
+	logger.Info("auth-daemon listening on https://127.0.0.1%s", s.addr)
+	if err := srv.ListenAndServeTLS("", ""); err != nil && err != http.ErrServerClosed {
+		return err
+	}
+	s.cleanupPrincipalsFile()
+	return nil
+}
+
+func (s *Server) cleanupPrincipalsFile() {
+	if s.cfg.PrincipalsFilePath != "" {
+		if err := os.Remove(s.cfg.PrincipalsFilePath); err != nil && !os.IsNotExist(err) {
+			logger.Warn("auth-daemon: remove principals file: %v", err)
+		}
+	}
+}

bind/shared_bind.go

@@ -144,6 +144,10 @@ type SharedBind struct {
 
 	// Callback for magic test responses (used for holepunch testing)
 	magicResponseCallback atomic.Pointer[func(addr netip.AddrPort, echoData []byte)]
+
+	// Rebinding state - used to keep receive goroutines alive during socket transition
+	rebinding     bool       // true when socket is being replaced
+	rebindingCond *sync.Cond // signaled when rebind completes
 }
 
 // MagicResponseCallback is the function signature for magic packet response callbacks
@@ -163,6 +167,9 @@ func New(udpConn *net.UDPConn) (*SharedBind, error) {
 		closeChan:       make(chan struct{}),
 	}
 
+	// Initialize the rebinding condition variable
+	bind.rebindingCond = sync.NewCond(&bind.mu)
+
 	// Initialize reference count to 1 (the creator holds the first reference)
 	bind.refCount.Store(1)
 
@@ -310,6 +317,109 @@ func (b *SharedBind) IsClosed() bool {
 	return b.closed.Load()
 }
 
+// GetPort returns the current UDP port the bind is using.
+// This is useful when rebinding to try to reuse the same port.
+func (b *SharedBind) GetPort() uint16 {
+	b.mu.RLock()
+	defer b.mu.RUnlock()
+	return b.port
+}
+
+// CloseSocket closes the underlying UDP connection to release the port,
+// but keeps the SharedBind in a state where it can accept a new connection via Rebind.
+// This allows the caller to close the old socket first, then bind a new socket
+// to the same port before calling Rebind.
+//
+// Returns the port that was being used, so the caller can attempt to rebind to it.
+// Sets the rebinding flag so receive goroutines will wait for the new socket
+// instead of exiting.
+func (b *SharedBind) CloseSocket() (uint16, error) {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+
+	if b.closed.Load() {
+		return 0, fmt.Errorf("bind is closed")
+	}
+
+	port := b.port
+
+	// Set rebinding flag BEFORE closing the socket so receive goroutines
+	// know to wait instead of exit
+	b.rebinding = true
+
+	// Close the old connection to release the port
+	if b.udpConn != nil {
+		logger.Debug("Closing UDP connection to release port %d (rebinding)", port)
+		b.udpConn.Close()
+		b.udpConn = nil
+	}
+
+	return port, nil
+}
+
+// Rebind replaces the underlying UDP connection with a new one.
+// This is necessary when network connectivity changes (e.g., WiFi to cellular
+// transition on macOS/iOS) and the old socket becomes stale.
+//
+// The caller is responsible for creating the new UDP connection and passing it here.
+// After rebind, the caller should trigger a hole punch to re-establish NAT mappings.
+//
+// Note: Call CloseSocket() first if you need to rebind to the same port, as the
+// old socket must be closed before a new socket can bind to the same port.
+func (b *SharedBind) Rebind(newConn *net.UDPConn) error {
+	if newConn == nil {
+		return fmt.Errorf("newConn cannot be nil")
+	}
+
+	b.mu.Lock()
+	defer b.mu.Unlock()
+
+	if b.closed.Load() {
+		return fmt.Errorf("bind is closed")
+	}
+
+	// Close the old connection if it's still open
+	// (it may have already been closed via CloseSocket)
+	if b.udpConn != nil {
+		logger.Debug("Closing old UDP connection during rebind")
+		b.udpConn.Close()
+	}
+
+	// Set up the new connection
+	b.udpConn = newConn
+
+	// Update packet connections for the new socket
+	if runtime.GOOS == "linux" || runtime.GOOS == "android" {
+		b.ipv4PC = ipv4.NewPacketConn(newConn)
+		b.ipv6PC = ipv6.NewPacketConn(newConn)
+
+		// Re-initialize message buffers for batch operations
+		batchSize := wgConn.IdealBatchSize
+		b.ipv4Msgs = make([]ipv4.Message, batchSize)
+		for i := range b.ipv4Msgs {
+			b.ipv4Msgs[i].OOB = make([]byte, 0)
+		}
+	} else {
+		// For non-Linux platforms, still set up ipv4PC for consistency
+		b.ipv4PC = ipv4.NewPacketConn(newConn)
+		b.ipv6PC = ipv6.NewPacketConn(newConn)
+	}
+
+	// Update the port
+	if addr, ok := newConn.LocalAddr().(*net.UDPAddr); ok {
+		b.port = uint16(addr.Port)
+		logger.Info("Rebound UDP socket to port %d", b.port)
+	}
+
+	// Clear the rebinding flag and wake up any waiting receive goroutines
+	b.rebinding = false
+	b.rebindingCond.Broadcast()
+
+	logger.Debug("Rebind complete, signaled waiting receive goroutines")
+
+	return nil
+}
+
 // SetMagicResponseCallback sets a callback function that will be called when
 // a magic test response packet is received. This is used for holepunch testing.
 // Pass nil to clear the callback.
@@ -392,24 +502,77 @@ func (b *SharedBind) Open(uport uint16) ([]wgConn.ReceiveFunc, uint16, error) {
 // makeReceiveSocket creates a receive function for physical UDP socket packets
 func (b *SharedBind) makeReceiveSocket() wgConn.ReceiveFunc {
 	return func(bufs [][]byte, sizes []int, eps []wgConn.Endpoint) (n int, err error) {
-		if b.closed.Load() {
-			return 0, net.ErrClosed
-		}
+		for {
+			if b.closed.Load() {
+				return 0, net.ErrClosed
+			}
 
-		b.mu.RLock()
-		conn := b.udpConn
-		pc := b.ipv4PC
-		b.mu.RUnlock()
+			b.mu.RLock()
+			conn := b.udpConn
+			pc := b.ipv4PC
+			b.mu.RUnlock()
 
-		if conn == nil {
-			return 0, net.ErrClosed
-		}
+			if conn == nil {
+				// Socket is nil - check if we're rebinding or truly closed
+				if b.closed.Load() {
+					return 0, net.ErrClosed
+				}
 
-		// Use batch reading on Linux for performance
-		if pc != nil && (runtime.GOOS == "linux" || runtime.GOOS == "android") {
-			return b.receiveIPv4Batch(pc, bufs, sizes, eps)
+				// Wait for rebind to complete
+				b.mu.Lock()
+				for b.rebinding && !b.closed.Load() {
+					logger.Debug("Receive goroutine waiting for socket rebind to complete")
+					b.rebindingCond.Wait()
+				}
+				b.mu.Unlock()
+
+				// Check again after waking up
+				if b.closed.Load() {
+					return 0, net.ErrClosed
+				}
+
+				// Loop back to retry with new socket
+				continue
+			}
+
+			// Use batch reading on Linux for performance
+			var n int
+			var err error
+			if pc != nil && (runtime.GOOS == "linux" || runtime.GOOS == "android") {
+				n, err = b.receiveIPv4Batch(pc, bufs, sizes, eps)
+			} else {
+				n, err = b.receiveIPv4Simple(conn, bufs, sizes, eps)
+			}
+
+			if err != nil {
+				// Check if this error is due to rebinding
+				b.mu.RLock()
+				rebinding := b.rebinding
+				b.mu.RUnlock()
+
+				if rebinding {
+					logger.Debug("Receive got error during rebind, waiting for new socket: %v", err)
+					// Wait for rebind to complete and retry
+					b.mu.Lock()
+					for b.rebinding && !b.closed.Load() {
+						b.rebindingCond.Wait()
+					}
+					b.mu.Unlock()
+
+					if b.closed.Load() {
+						return 0, net.ErrClosed
+					}
+
+					// Retry with new socket
+					continue
+				}
+
+				// Not rebinding, return the error
+				return 0, err
+			}
+
+			return n, nil
 		}
-		return b.receiveIPv4Simple(conn, bufs, sizes, eps)
 	}
 }
 
@@ -492,6 +655,8 @@ func (b *SharedBind) receiveIPv4Batch(pc *ipv4.PacketConn, bufs [][]byte, sizes
 
 // receiveIPv4Simple uses simple ReadFromUDP for non-Linux platforms
 func (b *SharedBind) receiveIPv4Simple(conn *net.UDPConn, bufs [][]byte, sizes []int, eps []wgConn.Endpoint) (int, error) {
+	// No read deadline - we rely on socket close to unblock during rebind.
+	// The caller (makeReceiveSocket) handles rebind state when errors occur.
 	for {
 		n, addr, err := conn.ReadFromUDP(bufs[0])
 		if err != nil {
@@ -523,7 +688,7 @@ func (b *SharedBind) receiveIPv4Simple(conn *net.UDPConn, bufs [][]byte, sizes [
 func (b *SharedBind) handleMagicPacket(data []byte, addr *net.UDPAddr) bool {
 	// Check if this is a test request packet
 	if len(data) >= MagicTestRequestLen && bytes.HasPrefix(data, MagicTestRequest) {
-		logger.Debug("Received magic test REQUEST from %s, sending response", addr.String())
+		// logger.Debug("Received magic test REQUEST from %s, sending response", addr.String())
 		// Extract the random data portion to echo back
 		echoData := data[len(MagicTestRequest) : len(MagicTestRequest)+MagicPacketDataLen]
 
@@ -546,7 +711,7 @@ func (b *SharedBind) handleMagicPacket(data []byte, addr *net.UDPAddr) bool {
 
 	// Check if this is a test response packet
 	if len(data) >= MagicTestResponseLen && bytes.HasPrefix(data, MagicTestResponse) {
-		logger.Debug("Received magic test RESPONSE from %s", addr.String())
+		// logger.Debug("Received magic test RESPONSE from %s", addr.String())
 		// Extract the echoed data
 		echoData := data[len(MagicTestResponse) : len(MagicTestResponse)+MagicPacketDataLen]
 

bind/shared_bind_test.go

@@ -553,53 +553,3 @@ func TestSocketRouting(t *testing.T) {
 		t.Errorf("Expected response from physical port %d, got %d", physicalAddr.Port, fromAddr.Port)
 	}
 }
-
-// TestClearNetstackConn tests that clearing the netstack connection works correctly
-func TestClearNetstackConn(t *testing.T) {
-	physicalConn, err := net.ListenUDP("udp", &net.UDPAddr{IP: net.IPv4zero, Port: 0})
-	if err != nil {
-		t.Fatalf("Failed to create physical UDP connection: %v", err)
-	}
-
-	sharedBind, err := New(physicalConn)
-	if err != nil {
-		t.Fatalf("Failed to create SharedBind: %v", err)
-	}
-	defer sharedBind.Close()
-
-	// Set a netstack connection
-	netstackConn, err := net.ListenUDP("udp", &net.UDPAddr{IP: net.IPv4zero, Port: 0})
-	if err != nil {
-		t.Fatalf("Failed to create netstack UDP connection: %v", err)
-	}
-	defer netstackConn.Close()
-
-	sharedBind.SetNetstackConn(netstackConn)
-
-	// Inject a packet to track an endpoint
-	testAddrPort := netip.MustParseAddrPort("192.168.1.100:51820")
-	err = sharedBind.InjectPacket([]byte("test"), testAddrPort)
-	if err != nil {
-		t.Fatalf("InjectPacket failed: %v", err)
-	}
-
-	// Verify the endpoint is tracked
-	_, tracked := sharedBind.netstackEndpoints.Load(testAddrPort.String())
-	if !tracked {
-		t.Error("Expected endpoint to be tracked as netstack-sourced")
-	}
-
-	// Clear the netstack connection
-	sharedBind.ClearNetstackConn()
-
-	// Verify the netstack connection is cleared
-	if sharedBind.GetNetstackConn() != nil {
-		t.Error("Expected netstack connection to be nil after clear")
-	}
-
-	// Verify the tracked endpoints are cleared
-	_, stillTracked := sharedBind.netstackEndpoints.Load(testAddrPort.String())
-	if stillTracked {
-		t.Error("Expected endpoint tracking to be cleared")
-	}
-}

clients.go

@@ -24,7 +24,7 @@ func setupClients(client *websocket.Client) {
 
 	host = strings.TrimSuffix(host, "/")
 
-	logger.Info("Setting up clients with netstack2...")
+	logger.Debug("Setting up clients with netstack2...")
 
 	// if useNativeInterface is true make sure we have permission to use native interface
 	if useNativeInterface {
@@ -37,7 +37,7 @@ func setupClients(client *websocket.Client) {
 	}
 
 	// Create WireGuard service
-	wgService, err = wgnetstack.NewWireGuardService(interfaceName, mtuInt, host, id, client, dns, useNativeInterface)
+	wgService, err = wgnetstack.NewWireGuardService(interfaceName, port, mtuInt, host, id, client, dns, useNativeInterface)
 	if err != nil {
 		logger.Fatal("Failed to create WireGuard service: %v", err)
 	}
@@ -63,7 +63,7 @@ func closeClients() {
 	}
 }
 
-func clientsHandleNewtConnection(publicKey string, endpoint string) {
+func clientsHandleNewtConnection(publicKey string, endpoint string, relayPort uint16) {
 	if !ready {
 		return
 	}
@@ -77,7 +77,7 @@ func clientsHandleNewtConnection(publicKey string, endpoint string) {
 	endpoint = strings.Join(parts[:len(parts)-1], ":")
 
 	if wgService != nil {
-		wgService.StartHolepunch(publicKey, endpoint)
+		wgService.StartHolepunch(publicKey, endpoint, relayPort)
 	}
 }
 

clients/clients.go

@@ -14,6 +14,7 @@ import (
 	"time"
 
 	"github.com/fosrl/newt/bind"
+	newtDevice "github.com/fosrl/newt/device"
 	"github.com/fosrl/newt/holepunch"
 	"github.com/fosrl/newt/logger"
 	"github.com/fosrl/newt/netstack2"
@@ -22,7 +23,6 @@ import (
 	"github.com/fosrl/newt/websocket"
 	"github.com/fosrl/newt/wgtester"
 	"golang.zx2c4.com/wireguard/device"
-	"golang.zx2c4.com/wireguard/ipc"
 	"golang.zx2c4.com/wireguard/tun"
 	"golang.zx2c4.com/wireguard/tun/netstack"
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
@@ -40,12 +40,14 @@ type Target struct {
 	SourcePrefix string      `json:"sourcePrefix"`
 	DestPrefix   string      `json:"destPrefix"`
 	RewriteTo    string      `json:"rewriteTo,omitempty"`
+	DisableIcmp  bool        `json:"disableIcmp,omitempty"`
 	PortRange    []PortRange `json:"portRange,omitempty"`
 }
 
 type PortRange struct {
-	Min uint16 `json:"min"`
-	Max uint16 `json:"max"`
+	Min      uint16 `json:"min"`
+	Max      uint16 `json:"max"`
+	Protocol string `json:"protocol"` // "tcp" or "udp"
 }
 
 type Peer struct {
@@ -104,17 +106,21 @@ type WireGuardService struct {
 	wgTesterServer     *wgtester.Server
 }
 
-func NewWireGuardService(interfaceName string, mtu int, host string, newtId string, wsClient *websocket.Client, dns string, useNativeInterface bool) (*WireGuardService, error) {
+func NewWireGuardService(interfaceName string, port uint16, mtu int, host string, newtId string, wsClient *websocket.Client, dns string, useNativeInterface bool) (*WireGuardService, error) {
 	key, err := wgtypes.GeneratePrivateKey()
 	if err != nil {
 		return nil, fmt.Errorf("failed to generate private key: %v", err)
 	}
 
-	// Find an available port
-	port, err := util.FindAvailableUDPPort(49152, 65535)
-
-	if err != nil {
-		return nil, fmt.Errorf("error finding available port: %v", err)
+	logger.Debug("+++++++++++++++++++++++++++++++= the port is %d", port)
+	
+	if port == 0 {
+		// Find an available port
+		portRandom, err := util.FindAvailableUDPPort(49152, 65535)
+		if err != nil {
+			return nil, fmt.Errorf("error finding available port: %v", err)
+		}
+		port = uint16(portRandom)
 	}
 
 	// Create shared UDP socket for both holepunch and WireGuard
@@ -137,7 +143,7 @@ func NewWireGuardService(interfaceName string, mtu int, host string, newtId stri
 	// Add a reference for the hole punch manager (creator already has one reference for WireGuard)
 	sharedBind.AddRef()
 
-	logger.Info("Created shared UDP socket on port %d (refcount: %d)", port, sharedBind.GetRefCount())
+	logger.Debug("Created shared UDP socket on port %d (refcount: %d)", port, sharedBind.GetRefCount())
 
 	// Parse DNS addresses
 	dnsAddrs := []netip.Addr{netip.MustParseAddr(dns)}
@@ -266,16 +272,21 @@ func (s *WireGuardService) SetOnNetstackClose(callback func()) {
 }
 
 // StartHolepunch starts hole punching to a specific endpoint
-func (s *WireGuardService) StartHolepunch(publicKey string, endpoint string) {
+func (s *WireGuardService) StartHolepunch(publicKey string, endpoint string, relayPort uint16) {
 	if s.holePunchManager == nil {
 		logger.Warn("Hole punch manager not initialized")
 		return
 	}
 
+	if relayPort == 0 {
+	    relayPort = 21820
+	}
+
 	// Convert websocket.ExitNode to holepunch.ExitNode
 	hpExitNodes := []holepunch.ExitNode{
 		{
 			Endpoint:  endpoint,
+			RelayPort: relayPort,
 			PublicKey: publicKey,
 		},
 	}
@@ -285,7 +296,7 @@ func (s *WireGuardService) StartHolepunch(publicKey string, endpoint string) {
 		logger.Warn("Failed to start hole punch: %v", err)
 	}
 
-	logger.Info("Starting hole punch to %s with public key: %s", endpoint, publicKey)
+	logger.Debug("Starting hole punch to %s with public key: %s", endpoint, publicKey)
 }
 
 // StartDirectUDPRelay starts a direct UDP relay from the main tunnel netstack to the clients' WireGuard.
@@ -332,7 +343,7 @@ func (s *WireGuardService) StartDirectUDPRelay(tunnelIP string) error {
 	// Set the netstack connection on the SharedBind so responses go back through the tunnel
 	s.sharedBind.SetNetstackConn(listener)
 
-	logger.Info("Started direct UDP relay on %s:%d (bidirectional via SharedBind)", tunnelIP, s.Port)
+	logger.Debug("Started direct UDP relay on %s:%d (bidirectional via SharedBind)", tunnelIP, s.Port)
 
 	// Start the relay goroutine to read from netstack and inject into SharedBind
 	s.directRelayWg.Add(1)
@@ -350,7 +361,7 @@ func (s *WireGuardService) runDirectUDPRelay(listener net.PacketConn) {
 	// Note: Don't close listener here - it's also used by SharedBind for sending responses
 	// It will be closed when the relay is stopped
 
-	logger.Info("Direct UDP relay started (bidirectional through SharedBind)")
+	logger.Debug("Direct UDP relay started (bidirectional through SharedBind)")
 
 	buf := make([]byte, 65535) // Max UDP packet size
 
@@ -436,7 +447,7 @@ func (s *WireGuardService) LoadRemoteConfig() error {
 		"port":      s.Port,
 	}, 2*time.Second)
 
-	logger.Info("Requesting WireGuard configuration from remote server")
+	logger.Debug("Requesting WireGuard configuration from remote server")
 	go s.periodicBandwidthCheck()
 
 	return nil
@@ -446,7 +457,7 @@ func (s *WireGuardService) handleConfig(msg websocket.WSMessage) {
 	var config WgConfig
 
 	logger.Debug("Received message: %v", msg)
-	logger.Info("Received WireGuard clients configuration from remote server")
+	logger.Debug("Received WireGuard clients configuration from remote server")
 
 	jsonData, err := json.Marshal(msg.Data)
 	if err != nil {
@@ -468,6 +479,8 @@ func (s *WireGuardService) handleConfig(msg websocket.WSMessage) {
 	// Ensure the WireGuard interface and peers are configured
 	if err := s.ensureWireguardInterface(config); err != nil {
 		logger.Error("Failed to ensure WireGuard interface: %v", err)
+		logger.Error("Clients functionality will be disabled until the interface can be created")
+		return
 	}
 
 	if err := s.ensureWireguardPeers(config.Peers); err != nil {
@@ -477,6 +490,8 @@ func (s *WireGuardService) handleConfig(msg websocket.WSMessage) {
 	if err := s.ensureTargets(config.Targets); err != nil {
 		logger.Error("Failed to ensure WireGuard targets: %v", err)
 	}
+
+	logger.Info("Client connectivity setup. Ready to accept connections from clients!")
 }
 
 func (s *WireGuardService) ensureWireguardInterface(wgconfig WgConfig) error {
@@ -522,17 +537,17 @@ func (s *WireGuardService) ensureWireguardInterface(wgconfig WgConfig) error {
 		// Create WireGuard device using the shared bind
 		s.device = device.NewDevice(s.tun, s.sharedBind, device.NewLogger(
 			device.LogLevelSilent,
-			"wireguard: ",
+			"client-wireguard: ",
 		))
 
 		fileUAPI, err := func() (*os.File, error) {
-			return ipc.UAPIOpen(interfaceName)
+			return newtDevice.UapiOpen(interfaceName)
 		}()
 		if err != nil {
 			logger.Error("UAPI listen error: %v", err)
 		}
 
-		uapiListener, err := ipc.UAPIListen(interfaceName, fileUAPI)
+		uapiListener, err := newtDevice.UapiListen(interfaceName, fileUAPI)
 		if err != nil {
 			logger.Error("Failed to listen on uapi socket: %v", err)
 			os.Exit(1)
@@ -590,8 +605,9 @@ func (s *WireGuardService) ensureWireguardInterface(wgconfig WgConfig) error {
 		s.dns,
 		s.mtu,
 		netstack2.NetTunOptions{
-			EnableTCPProxy: true,
-			EnableUDPProxy: true,
+			EnableTCPProxy:  true,
+			EnableUDPProxy:  true,
+			EnableICMPProxy: true,
 		},
 	)
 	if err != nil {
@@ -623,7 +639,7 @@ func (s *WireGuardService) ensureWireguardInterface(wgconfig WgConfig) error {
 		return fmt.Errorf("failed to bring up WireGuard device: %v", err)
 	}
 
-	logger.Info("WireGuard netstack device created and configured")
+	logger.Debug("WireGuard netstack device created and configured")
 
 	// Release the mutex before calling the callback
 	s.mu.Unlock()
@@ -642,6 +658,11 @@ func (s *WireGuardService) ensureWireguardPeers(peers []Peer) error {
 	// For netstack, we need to manage peers differently
 	// We'll configure peers directly on the device using IPC
 
+	// Check if device is initialized
+	if s.device == nil {
+		return fmt.Errorf("WireGuard device is not initialized")
+	}
+
 	// First, clear all existing peers by getting current config and removing them
 	currentConfig, err := s.device.IpcGet()
 	if err != nil {
@@ -697,14 +718,15 @@ func (s *WireGuardService) ensureTargets(targets []Target) error {
 		var portRanges []netstack2.PortRange
 		for _, pr := range target.PortRange {
 			portRanges = append(portRanges, netstack2.PortRange{
-				Min: pr.Min,
-				Max: pr.Max,
+				Min:      pr.Min,
+				Max:      pr.Max,
+				Protocol: pr.Protocol,
 			})
 		}
 
-		s.tnet.AddProxySubnetRule(sourcePrefix, destPrefix, target.RewriteTo, portRanges)
+		s.tnet.AddProxySubnetRule(sourcePrefix, destPrefix, target.RewriteTo, portRanges, target.DisableIcmp)
 
-		logger.Info("Added target subnet from %s to %s rewrite to %s with port ranges: %v", target.SourcePrefix, target.DestPrefix, target.RewriteTo, target.PortRange)
+		logger.Info("Added target subnet from %s to %s rewrite to %s with port ranges: %v disableIcmp: %v", target.SourcePrefix, target.DestPrefix, target.RewriteTo, target.PortRange, target.DisableIcmp)
 	}
 
 	return nil
@@ -1015,21 +1037,22 @@ func (s *WireGuardService) processPeerBandwidth(publicKey string, rxBytes, txByt
 			// Update the last reading
 			s.lastReadings[publicKey] = currentReading
 
-			return &PeerBandwidth{
-				PublicKey: publicKey,
-				BytesIn:   bytesInMB,
-				BytesOut:  bytesOutMB,
+			// Only return bandwidth data if there was an increase
+			if bytesInDiff > 0 || bytesOutDiff > 0 {
+				return &PeerBandwidth{
+					PublicKey: publicKey,
+					BytesIn:   bytesInMB,
+					BytesOut:  bytesOutMB,
+				}
 			}
+			
+			return nil
 		}
 	}
 
-	// For first reading or if readings are too close together, report 0
+	// For first reading or if readings are too close together, don't report
 	s.lastReadings[publicKey] = currentReading
-	return &PeerBandwidth{
-		PublicKey: publicKey,
-		BytesIn:   0,
-		BytesOut:  0,
-	}
+	return nil
 }
 
 func (s *WireGuardService) reportPeerBandwidth() error {
@@ -1090,12 +1113,13 @@ func (s *WireGuardService) handleAddTarget(msg websocket.WSMessage) {
 			portRanges = append(portRanges, netstack2.PortRange{
 				Min: pr.Min,
 				Max: pr.Max,
+				Protocol:    pr.Protocol,
 			})
 		}
 
-		s.tnet.AddProxySubnetRule(sourcePrefix, destPrefix, target.RewriteTo, portRanges)
+		s.tnet.AddProxySubnetRule(sourcePrefix, destPrefix, target.RewriteTo, portRanges, target.DisableIcmp)
 
-		logger.Info("Added target subnet from %s to %s rewrite to %s with port ranges: %v", target.SourcePrefix, target.DestPrefix, target.RewriteTo, target.PortRange)
+		logger.Info("Added target subnet from %s to %s rewrite to %s with port ranges: %v disableIcmp: %v", target.SourcePrefix, target.DestPrefix, target.RewriteTo, target.PortRange, target.DisableIcmp)
 	}
 }
 
@@ -1205,13 +1229,14 @@ func (s *WireGuardService) handleUpdateTarget(msg websocket.WSMessage) {
 		var portRanges []netstack2.PortRange
 		for _, pr := range target.PortRange {
 			portRanges = append(portRanges, netstack2.PortRange{
-				Min: pr.Min,
-				Max: pr.Max,
+				Min:         pr.Min,
+				Max:         pr.Max,
+				Protocol:    pr.Protocol,
 			})
 		}
 
-		s.tnet.AddProxySubnetRule(sourcePrefix, destPrefix, target.RewriteTo, portRanges)
-		logger.Info("Added target subnet from %s to %s rewrite to %s with port ranges: %v", target.SourcePrefix, target.DestPrefix, target.RewriteTo, target.PortRange)
+		s.tnet.AddProxySubnetRule(sourcePrefix, destPrefix, target.RewriteTo, portRanges, target.DisableIcmp)
+		logger.Info("Added target subnet from %s to %s rewrite to %s with port ranges: %v disableIcmp: %v", target.SourcePrefix, target.DestPrefix, target.RewriteTo, target.PortRange, target.DisableIcmp)
 	}
 }
 

clients/permissions/permissions_android.go

@@ -0,0 +1,8 @@
+//go:build android
+
+package permissions
+
+// CheckNativeInterfacePermissions always allows permission on Android.
+func CheckNativeInterfacePermissions() error {
+	return nil
+}

clients/permissions/permissions_darwin.go

@@ -1,4 +1,4 @@
-//go:build darwin
+//go:build darwin && !ios
 
 package permissions
 

clients/permissions/permissions_freebsd.go

@@ -0,0 +1,57 @@
+//go:build freebsd
+
+package permissions
+
+import (
+	"fmt"
+	"os"
+
+	"github.com/fosrl/newt/logger"
+)
+
+const (
+	// TUN device on FreeBSD
+	tunDevice = "/dev/tun"
+	ifnamsiz  = 16
+	iffTun    = 0x0001
+	iffNoPi   = 0x1000
+)
+
+// ifReq is the structure for TUN interface configuration
+type ifReq struct {
+	Name  [ifnamsiz]byte
+	Flags uint16
+	_     [22]byte // padding to match kernel structure
+}
+
+// CheckNativeInterfacePermissions checks if the process has sufficient
+// permissions to create a native TUN interface on FreeBSD.
+// This requires root privileges (UID 0).
+func CheckNativeInterfacePermissions() error {
+	logger.Debug("Checking native interface permissions on FreeBSD")
+
+	// Check if running as root
+	if os.Geteuid() == 0 {
+		logger.Debug("Running as root, sufficient permissions for native TUN interface")
+		return nil
+	}
+
+	// On FreeBSD, only root can create TUN interfaces
+	// Try to open the TUN device to verify
+	return tryOpenTunDevice()
+}
+
+// tryOpenTunDevice attempts to open the TUN device to verify permissions.
+// On FreeBSD, /dev/tun is a cloning device that creates a new interface
+// when opened.
+func tryOpenTunDevice() error {
+	// Try opening /dev/tun (cloning device)
+	f, err := os.OpenFile(tunDevice, os.O_RDWR, 0)
+	if err != nil {
+		return fmt.Errorf("cannot open %s: %v (need root privileges)", tunDevice, err)
+	}
+	defer f.Close()
+
+	logger.Debug("Successfully opened TUN device, sufficient permissions for native TUN interface")
+	return nil
+}

clients/permissions/permissions_ios.go

@@ -0,0 +1,8 @@
+//go:build ios
+
+package permissions
+
+// CheckNativeInterfacePermissions always allows permission on iOS.
+func CheckNativeInterfacePermissions() error {
+	return nil
+}

clients/permissions/permissions_linux.go

@@ -1,4 +1,4 @@
-//go:build linux
+//go:build linux && !android
 
 package permissions
 

clients/permissions/permissions_windows.go

@@ -6,14 +6,24 @@ import (
 	"fmt"
 
 	"golang.org/x/sys/windows"
+	"golang.org/x/sys/windows/svc"
 )
 
 // CheckNativeInterfacePermissions checks if the process has sufficient
 // permissions to create a native TUN interface on Windows.
-// This requires Administrator privileges.
+// This requires Administrator privileges and must be running as a Windows service.
 func CheckNativeInterfacePermissions() error {
+	// Check if running as a Windows service
+	isService, err := svc.IsWindowsService()
+	if err != nil {
+		return fmt.Errorf("failed to check if running as Windows service: %v", err)
+	}
+	if !isService {
+		return fmt.Errorf("native TUN interface requires running as a Windows service")
+	}
+
 	var sid *windows.SID
-	err := windows.AllocateAndInitializeSid(
+	err = windows.AllocateAndInitializeSid(
 		&windows.SECURITY_NT_AUTHORITY,
 		2,
 		windows.SECURITY_BUILTIN_DOMAIN_RID,

device/tun_unix.go

@@ -0,0 +1,44 @@
+//go:build !windows
+
+package device
+
+import (
+	"net"
+	"os"
+
+	"github.com/fosrl/newt/logger"
+	"golang.org/x/sys/unix"
+	"golang.zx2c4.com/wireguard/ipc"
+	"golang.zx2c4.com/wireguard/tun"
+)
+
+func CreateTUNFromFD(tunFd uint32, mtuInt int) (tun.Device, error) {
+	dupTunFd, err := unix.Dup(int(tunFd))
+	if err != nil {
+		logger.Error("Unable to dup tun fd: %v", err)
+		return nil, err
+	}
+
+	err = unix.SetNonblock(dupTunFd, true)
+	if err != nil {
+		unix.Close(dupTunFd)
+		return nil, err
+	}
+
+	file := os.NewFile(uintptr(dupTunFd), "/dev/tun")
+	device, err := tun.CreateTUNFromFile(file, mtuInt)
+	if err != nil {
+		file.Close()
+		return nil, err
+	}
+
+	return device, nil
+}
+
+func UapiOpen(interfaceName string) (*os.File, error) {
+	return ipc.UAPIOpen(interfaceName)
+}
+
+func UapiListen(interfaceName string, fileUAPI *os.File) (net.Listener, error) {
+	return ipc.UAPIListen(interfaceName, fileUAPI)
+}

device/tun_windows.go

@@ -0,0 +1,25 @@
+//go:build windows
+
+package device
+
+import (
+	"errors"
+	"net"
+	"os"
+
+	"golang.zx2c4.com/wireguard/ipc"
+	"golang.zx2c4.com/wireguard/tun"
+)
+
+func CreateTUNFromFD(tunFd uint32, mtuInt int) (tun.Device, error) {
+	return nil, errors.New("CreateTUNFromFile not supported on Windows")
+}
+
+func UapiOpen(interfaceName string) (*os.File, error) {
+	return nil, nil
+}
+
+func UapiListen(interfaceName string, fileUAPI *os.File) (net.Listener, error) {
+	// On Windows, UAPIListen only takes one parameter
+	return ipc.UAPIListen(interfaceName)
+}

docker-compose.yml

@@ -4,7 +4,7 @@ services:
     container_name: newt
     restart: unless-stopped
     environment:
-      - PANGOLIN_ENDPOINT=https://example.com
+      - PANGOLIN_ENDPOINT=https://app.pangolin.net
       - NEWT_ID=2ix2t8xk22ubpfy 
       - NEWT_SECRET=nnisrfsdfc7prqsp9ewo1dvtvci50j5uiqotez00dgap0ii2 
       - LOG_LEVEL=DEBUG

flake.lock

@@ -2,16 +2,16 @@
   "nodes": {
     "nixpkgs": {
       "locked": {
-        "lastModified": 1756217674,
-        "narHash": "sha256-TH1SfSP523QI7kcPiNtMAEuwZR3Jdz0MCDXPs7TS8uo=",
+        "lastModified": 1763934636,
+        "narHash": "sha256-9glbI7f1uU+yzQCq5LwLgdZqx6svOhZWkd4JRY265fc=",
         "owner": "NixOS",
         "repo": "nixpkgs",
-        "rev": "4e7667a90c167f7a81d906e5a75cba4ad8bee620",
+        "rev": "ee09932cedcef15aaf476f9343d1dea2cb77e261",
         "type": "github"
       },
       "original": {
         "owner": "NixOS",
-        "ref": "nixos-25.05",
+        "ref": "nixpkgs-unstable",
         "repo": "nixpkgs",
         "type": "github"
       }

flake.nix

@@ -2,7 +2,7 @@
   description = "newt - A tunneling client for Pangolin";
 
   inputs = {
-    nixpkgs.url = "github:NixOS/nixpkgs/nixos-25.05";
+    nixpkgs.url = "github:NixOS/nixpkgs/nixpkgs-unstable";
   };
 
   outputs =
@@ -22,30 +22,49 @@
         system:
         let
           pkgs = pkgsFor system;
+          inherit (pkgs) lib;
 
           # Update version when releasing
-          version = "1.4.2";
-
-          # Update the version in a new source tree
-          srcWithReplacedVersion = pkgs.runCommand "newt-src-with-version" { } ''
-            cp -r ${./.} $out
-            chmod -R +w $out
-            rm -rf $out/.git $out/result $out/.envrc $out/.direnv
-            sed -i "s/version_replaceme/${version}/g" $out/main.go
-          '';
+          version = "1.8.0";
         in
         {
           default = self.packages.${system}.pangolin-newt;
+
           pangolin-newt = pkgs.buildGoModule {
             pname = "pangolin-newt";
-            version = version;
-            src = srcWithReplacedVersion;
-            vendorHash = "sha256-PENsCO2yFxLVZNPgx2OP+gWVNfjJAfXkwWS7tzlm490=";
-            meta = with pkgs.lib; {
+            inherit version;
+            src = pkgs.nix-gitignore.gitignoreSource [ ] ./.;
+
+            vendorHash = "sha256-vy6Dqjek7pLdASbCrM9snq5Dt9lbwNJ0AuQboy1JWNQ=";
+
+            nativeInstallCheckInputs = [ pkgs.versionCheckHook ];
+
+            env = {
+              CGO_ENABLED = 0;
+            };
+
+            ldflags = [
+              "-s"
+              "-w"
+              "-X main.newtVersion=${version}"
+            ];
+
+            # Tests are broken due to a lack of Internet.
+            # Disable running `go test`, and instead do
+            # a simple version check instead.
+            doCheck = false;
+            doInstallCheck = true;
+
+            versionCheckProgramArg = [ "-version" ];
+
+            meta = {
               description = "A tunneling client for Pangolin";
               homepage = "https://github.com/fosrl/newt";
-              license = licenses.gpl3;
-              maintainers = [ ];
+              license = lib.licenses.gpl3;
+              maintainers = [
+                lib.maintainers.water-sucks
+              ];
+              mainProgram = "newt";
             };
           };
         }
@@ -54,10 +73,20 @@
         system:
         let
           pkgs = pkgsFor system;
+
+          inherit (pkgs)
+            go
+            gopls
+            gotools
+            go-outline
+            gopkgs
+            godef
+            golint
+            ;
         in
         {
           default = pkgs.mkShell {
-            buildInputs = with pkgs; [
+            buildInputs = [
               go
               gopls
               gotools

go.mod

@@ -1,31 +1,33 @@
 module github.com/fosrl/newt
 
-go 1.25
+go 1.25.0
 
 require (
-	github.com/docker/docker v28.5.1+incompatible
-	github.com/google/gopacket v1.1.19
+	github.com/docker/docker v28.5.2+incompatible
+	github.com/gaissmai/bart v0.26.0
 	github.com/gorilla/websocket v1.5.3
 	github.com/prometheus/client_golang v1.23.2
 	github.com/vishvananda/netlink v1.3.1
-	go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.63.0
-	go.opentelemetry.io/contrib/instrumentation/runtime v0.63.0
-	go.opentelemetry.io/otel v1.38.0
-	go.opentelemetry.io/otel/exporters/otlp/otlpmetric/otlpmetricgrpc v1.38.0
-	go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc v1.38.0
-	go.opentelemetry.io/otel/exporters/prometheus v0.60.0
-	go.opentelemetry.io/otel/metric v1.38.0
-	go.opentelemetry.io/otel/sdk v1.38.0
-	go.opentelemetry.io/otel/sdk/metric v1.38.0
-	golang.org/x/crypto v0.44.0
+	go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.66.0
+	go.opentelemetry.io/contrib/instrumentation/runtime v0.66.0
+	go.opentelemetry.io/otel v1.41.0
+	go.opentelemetry.io/otel/exporters/otlp/otlpmetric/otlpmetricgrpc v1.41.0
+	go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc v1.41.0
+	go.opentelemetry.io/otel/exporters/prometheus v0.63.0
+	go.opentelemetry.io/otel/metric v1.41.0
+	go.opentelemetry.io/otel/sdk v1.41.0
+	go.opentelemetry.io/otel/sdk/metric v1.41.0
+	golang.org/x/crypto v0.48.0
 	golang.org/x/exp v0.0.0-20251113190631-e25ba8c21ef6
-	golang.org/x/net v0.47.0
+	golang.org/x/net v0.51.0
+	golang.org/x/sys v0.41.0
 	golang.zx2c4.com/wireguard v0.0.0-20250521234502-f333402bd9cb
 	golang.zx2c4.com/wireguard/wgctrl v0.0.0-20241231184526-a9ab2273dd10
-	google.golang.org/grpc v1.76.0
+	golang.zx2c4.com/wireguard/windows v0.5.3
+	google.golang.org/grpc v1.79.1
 	gopkg.in/yaml.v3 v3.0.1
 	gvisor.dev/gvisor v0.0.0-20250503011706-39ed1f5ac29c
-	software.sslmate.com/src/go-pkcs12 v0.6.0
+	software.sslmate.com/src/go-pkcs12 v0.7.0
 )
 
 require (
@@ -42,14 +44,8 @@ require (
 	github.com/go-logr/logr v1.4.3 // indirect
 	github.com/go-logr/stdr v1.2.2 // indirect
 	github.com/google/btree v1.1.3 // indirect
-	github.com/google/go-cmp v0.7.0 // indirect
 	github.com/google/uuid v1.6.0 // indirect
-	github.com/grafana/regexp v0.0.0-20240518133315-a468a5bfb3bc // indirect
-	github.com/grpc-ecosystem/grpc-gateway/v2 v2.27.2 // indirect
-	github.com/josharian/native v1.1.0 // indirect
-	github.com/mdlayher/genetlink v1.3.2 // indirect
-	github.com/mdlayher/netlink v1.7.2 // indirect
-	github.com/mdlayher/socket v0.5.1 // indirect
+	github.com/grpc-ecosystem/grpc-gateway/v2 v2.28.0 // indirect
 	github.com/moby/docker-image-spec v1.3.1 // indirect
 	github.com/moby/sys/atomicwriter v0.1.0 // indirect
 	github.com/moby/term v0.5.2 // indirect
@@ -59,24 +55,23 @@ require (
 	github.com/opencontainers/image-spec v1.1.0 // indirect
 	github.com/pkg/errors v0.9.1 // indirect
 	github.com/prometheus/client_model v0.6.2 // indirect
-	github.com/prometheus/common v0.66.1 // indirect
-	github.com/prometheus/otlptranslator v0.0.2 // indirect
-	github.com/prometheus/procfs v0.17.0 // indirect
+	github.com/prometheus/common v0.67.5 // indirect
+	github.com/prometheus/otlptranslator v1.0.0 // indirect
+	github.com/prometheus/procfs v0.19.2 // indirect
 	github.com/vishvananda/netns v0.0.5 // indirect
-	go.opentelemetry.io/auto/sdk v1.1.0 // indirect
-	go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.38.0 // indirect
+	go.opentelemetry.io/auto/sdk v1.2.1 // indirect
+	go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.41.0 // indirect
 	go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp v1.38.0 // indirect
-	go.opentelemetry.io/otel/trace v1.38.0 // indirect
-	go.opentelemetry.io/proto/otlp v1.7.1 // indirect
-	go.yaml.in/yaml/v2 v2.4.2 // indirect
-	golang.org/x/mod v0.30.0 // indirect
-	golang.org/x/sync v0.18.0 // indirect
-	golang.org/x/sys v0.38.0 // indirect
-	golang.org/x/text v0.31.0 // indirect
+	go.opentelemetry.io/otel/trace v1.41.0 // indirect
+	go.opentelemetry.io/proto/otlp v1.9.0 // indirect
+	go.yaml.in/yaml/v2 v2.4.3 // indirect
+	golang.org/x/mod v0.32.0 // indirect
+	golang.org/x/sync v0.19.0 // indirect
+	golang.org/x/text v0.34.0 // indirect
 	golang.org/x/time v0.12.0 // indirect
-	golang.org/x/tools v0.39.0 // indirect
+	golang.org/x/tools v0.41.0 // indirect
 	golang.zx2c4.com/wintun v0.0.0-20230126152724-0fa3db229ce2 // indirect
-	google.golang.org/genproto/googleapis/api v0.0.0-20250825161204-c5933d9347a5 // indirect
-	google.golang.org/genproto/googleapis/rpc v0.0.0-20250825161204-c5933d9347a5 // indirect
-	google.golang.org/protobuf v1.36.8 // indirect
+	google.golang.org/genproto/googleapis/api v0.0.0-20260209200024-4cfbd4190f57 // indirect
+	google.golang.org/genproto/googleapis/rpc v0.0.0-20260209200024-4cfbd4190f57 // indirect
+	google.golang.org/protobuf v1.36.11 // indirect
 )

go.sum

@@ -18,14 +18,16 @@ github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/distribution/reference v0.6.0 h1:0IXCQ5g4/QMHHkarYzh5l+u8T3t73zM5QvfrDyIgxBk=
 github.com/distribution/reference v0.6.0/go.mod h1:BbU0aIcezP1/5jX/8MP0YiH4SdvB5Y4f/wlDRiLyi3E=
-github.com/docker/docker v28.5.1+incompatible h1:Bm8DchhSD2J6PsFzxC35TZo4TLGR2PdW/E69rU45NhM=
-github.com/docker/docker v28.5.1+incompatible/go.mod h1:eEKB0N0r5NX/I1kEveEz05bcu8tLC/8azJZsviup8Sk=
+github.com/docker/docker v28.5.2+incompatible h1:DBX0Y0zAjZbSrm1uzOkdr1onVghKaftjlSWt4AFexzM=
+github.com/docker/docker v28.5.2+incompatible/go.mod h1:eEKB0N0r5NX/I1kEveEz05bcu8tLC/8azJZsviup8Sk=
 github.com/docker/go-connections v0.6.0 h1:LlMG9azAe1TqfR7sO+NJttz1gy6KO7VJBh+pMmjSD94=
 github.com/docker/go-connections v0.6.0/go.mod h1:AahvXYshr6JgfUJGdDCs2b5EZG/vmaMAntpSFH5BFKE=
 github.com/docker/go-units v0.4.0 h1:3uh0PgVws3nIA0Q+MwDC8yjEPf9zjRfZZWXZYDct3Tw=
 github.com/docker/go-units v0.4.0/go.mod h1:fgPhTUdO+D/Jk86RDLlptpiXQzgHJF7gydDDbaIK4Dk=
 github.com/felixge/httpsnoop v1.0.4 h1:NFTV2Zj1bL4mc9sqWACXbQFVBBg2W3GPvqp8/ESS2Wg=
 github.com/felixge/httpsnoop v1.0.4/go.mod h1:m8KPJKqk1gH5J9DgRY2ASl2lWCfGKXixSwevea8zH2U=
+github.com/gaissmai/bart v0.26.0 h1:xOZ57E9hJLBiQaSyeZa9wgWhGuzfGACgqp4BE77OkO0=
+github.com/gaissmai/bart v0.26.0/go.mod h1:GREWQfTLRWz/c5FTOsIw+KkscuFkIV5t8Rp7Nd1Td5c=
 github.com/go-logr/logr v1.2.2/go.mod h1:jdQByPbusPIv2/zmleS9BjJVeZ6kBagPoEUsqbVz/1A=
 github.com/go-logr/logr v1.4.3 h1:CjnDlHq8ikf6E492q6eKboGOC0T8CDaOvkHCIg8idEI=
 github.com/go-logr/logr v1.4.3/go.mod h1:9T104GzyrTigFIr8wt5mBrctHMim0Nb2HLGrmQ40KvY=
@@ -37,18 +39,12 @@ github.com/google/btree v1.1.3 h1:CVpQJjYgC4VbzxeGVHfvZrv1ctoYCAI8vbl07Fcxlyg=
 github.com/google/btree v1.1.3/go.mod h1:qOPhT0dTNdNzV6Z/lhRX0YXUafgPLFUh+gZMl761Gm4=
 github.com/google/go-cmp v0.7.0 h1:wk8382ETsv4JYUZwIsn6YpYiWiBsYLSJiTsyBybVuN8=
 github.com/google/go-cmp v0.7.0/go.mod h1:pXiqmnSA92OHEEa9HXL2W4E7lf9JzCmGVUdgjX3N/iU=
-github.com/google/gopacket v1.1.19 h1:ves8RnFZPGiFnTS0uPQStjwru6uO6h+nlr9j6fL7kF8=
-github.com/google/gopacket v1.1.19/go.mod h1:iJ8V8n6KS+z2U1A8pUwu8bW5SyEMkXJB8Yo/Vo+TKTo=
 github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
 github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/gorilla/websocket v1.5.3 h1:saDtZ6Pbx/0u+bgYQ3q96pZgCzfhKXGPqt7kZ72aNNg=
 github.com/gorilla/websocket v1.5.3/go.mod h1:YR8l580nyteQvAITg2hZ9XVh4b55+EU/adAjf1fMHhE=
-github.com/grafana/regexp v0.0.0-20240518133315-a468a5bfb3bc h1:GN2Lv3MGO7AS6PrRoT6yV5+wkrOpcszoIsO4+4ds248=
-github.com/grafana/regexp v0.0.0-20240518133315-a468a5bfb3bc/go.mod h1:+JKpmjMGhpgPL+rXZ5nsZieVzvarn86asRlBg4uNGnk=
-github.com/grpc-ecosystem/grpc-gateway/v2 v2.27.2 h1:8Tjv8EJ+pM1xP8mK6egEbD1OgnVTyacbefKhmbLhIhU=
-github.com/grpc-ecosystem/grpc-gateway/v2 v2.27.2/go.mod h1:pkJQ2tZHJ0aFOVEEot6oZmaVEZcRme73eIFmhiVuRWs=
-github.com/josharian/native v1.1.0 h1:uuaP0hAbW7Y4l0ZRQ6C9zfb7Mg1mbFKry/xzDAfmtLA=
-github.com/josharian/native v1.1.0/go.mod h1:7X/raswPFr05uY3HiLlYeyQntB6OO7E/d2Cu7qoaN2w=
+github.com/grpc-ecosystem/grpc-gateway/v2 v2.28.0 h1:HWRh5R2+9EifMyIHV7ZV+MIZqgz+PMpZ14Jynv3O2Zs=
+github.com/grpc-ecosystem/grpc-gateway/v2 v2.28.0/go.mod h1:JfhWUomR1baixubs02l85lZYYOm7LV6om4ceouMv45c=
 github.com/klauspost/compress v1.18.0 h1:c/Cqfb0r+Yi+JtIEq73FWXVkRonBlf0CRNYc8Zttxdo=
 github.com/klauspost/compress v1.18.0/go.mod h1:2Pp+KzxcywXVXMr50+X0Q/Lsb43OQHYWRCY2AiWywWQ=
 github.com/kr/pretty v0.3.1 h1:flRD4NNwYAUpkphVc1HcthR4KEIFJ65n8Mw5qdRn3LE=
@@ -57,14 +53,6 @@ github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
 github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/kylelemons/godebug v1.1.0 h1:RPNrshWIDI6G2gRW9EHilWtl7Z6Sb1BR0xunSBf0SNc=
 github.com/kylelemons/godebug v1.1.0/go.mod h1:9/0rRGxNHcop5bhtWyNeEfOS8JIWk580+fNqagV/RAw=
-github.com/mdlayher/genetlink v1.3.2 h1:KdrNKe+CTu+IbZnm/GVUMXSqBBLqcGpRDa0xkQy56gw=
-github.com/mdlayher/genetlink v1.3.2/go.mod h1:tcC3pkCrPUGIKKsCsp0B3AdaaKuHtaxoJRz3cc+528o=
-github.com/mdlayher/netlink v1.7.2 h1:/UtM3ofJap7Vl4QWCPDGXY8d3GIY2UGSDbK+QWmY8/g=
-github.com/mdlayher/netlink v1.7.2/go.mod h1:xraEF7uJbxLhc5fpHL4cPe221LI2bdttWlU+ZGLfQSw=
-github.com/mdlayher/socket v0.5.1 h1:VZaqt6RkGkt2OE9l3GcC6nZkqD3xKeQLyfleW/uBcos=
-github.com/mdlayher/socket v0.5.1/go.mod h1:TjPLHI1UgwEv5J1B5q0zTZq12A/6H7nKmtTanQE37IQ=
-github.com/mikioh/ipaddr v0.0.0-20190404000644-d465c8ab6721 h1:RlZweED6sbSArvlE924+mUcZuXKLBHA35U7LN621Bws=
-github.com/mikioh/ipaddr v0.0.0-20190404000644-d465c8ab6721/go.mod h1:Ickgr2WtCLZ2MDGd4Gr0geeCH5HybhRJbonOgQpvSxc=
 github.com/moby/docker-image-spec v1.3.1 h1:jMKff3w6PgbfSa69GfNg+zN/XLhfXJGnEx3Nl2EsFP0=
 github.com/moby/docker-image-spec v1.3.1/go.mod h1:eKmb5VW8vQEh/BAr2yvVNvuiJuY6UIocYsFu/DxxRpo=
 github.com/moby/sys/atomicwriter v0.1.0 h1:kw5D/EqkBwsBFi0ss9v1VG3wIkVhzGvLklJ+w3A14Sw=
@@ -89,14 +77,14 @@ github.com/prometheus/client_golang v1.23.2 h1:Je96obch5RDVy3FDMndoUsjAhG5Edi49h
 github.com/prometheus/client_golang v1.23.2/go.mod h1:Tb1a6LWHB3/SPIzCoaDXI4I8UHKeFTEQ1YCr+0Gyqmg=
 github.com/prometheus/client_model v0.6.2 h1:oBsgwpGs7iVziMvrGhE53c/GrLUsZdHnqNwqPLxwZyk=
 github.com/prometheus/client_model v0.6.2/go.mod h1:y3m2F6Gdpfy6Ut/GBsUqTWZqCUvMVzSfMLjcu6wAwpE=
-github.com/prometheus/common v0.66.1 h1:h5E0h5/Y8niHc5DlaLlWLArTQI7tMrsfQjHV+d9ZoGs=
-github.com/prometheus/common v0.66.1/go.mod h1:gcaUsgf3KfRSwHY4dIMXLPV0K/Wg1oZ8+SbZk/HH/dA=
-github.com/prometheus/otlptranslator v0.0.2 h1:+1CdeLVrRQ6Psmhnobldo0kTp96Rj80DRXRd5OSnMEQ=
-github.com/prometheus/otlptranslator v0.0.2/go.mod h1:P8AwMgdD7XEr6QRUJ2QWLpiAZTgTE2UYgjlu3svompI=
-github.com/prometheus/procfs v0.17.0 h1:FuLQ+05u4ZI+SS/w9+BWEM2TXiHKsUQ9TADiRH7DuK0=
-github.com/prometheus/procfs v0.17.0/go.mod h1:oPQLaDAMRbA+u8H5Pbfq+dl3VDAvHxMUOVhe0wYB2zw=
-github.com/rogpeppe/go-internal v1.13.1 h1:KvO1DLK/DRN07sQ1LQKScxyZJuNnedQ5/wKSR38lUII=
-github.com/rogpeppe/go-internal v1.13.1/go.mod h1:uMEvuHeurkdAXX61udpOXGD/AzZDWNMNyH2VO9fmH0o=
+github.com/prometheus/common v0.67.5 h1:pIgK94WWlQt1WLwAC5j2ynLaBRDiinoAb86HZHTUGI4=
+github.com/prometheus/common v0.67.5/go.mod h1:SjE/0MzDEEAyrdr5Gqc6G+sXI67maCxzaT3A2+HqjUw=
+github.com/prometheus/otlptranslator v1.0.0 h1:s0LJW/iN9dkIH+EnhiD3BlkkP5QVIUVEoIwkU+A6qos=
+github.com/prometheus/otlptranslator v1.0.0/go.mod h1:vRYWnXvI6aWGpsdY/mOT/cbeVRBlPWtBNDb7kGR3uKM=
+github.com/prometheus/procfs v0.19.2 h1:zUMhqEW66Ex7OXIiDkll3tl9a1ZdilUOd/F6ZXw4Vws=
+github.com/prometheus/procfs v0.19.2/go.mod h1:M0aotyiemPhBCM0z5w87kL22CxfcH05ZpYlu+b4J7mw=
+github.com/rogpeppe/go-internal v1.14.1 h1:UQB4HGPB6osV0SQTLymcB4TgvyWu6ZyliaW0tI/otEQ=
+github.com/rogpeppe/go-internal v1.14.1/go.mod h1:MaRKkUm5W0goXpeCfT7UZI6fk/L7L7so1lCWt35ZSgc=
 github.com/sirupsen/logrus v1.9.3 h1:dueUQJ1C2q9oE3F7wvmSGAaVtTmUizReu6fjN8uqzbQ=
 github.com/sirupsen/logrus v1.9.3/go.mod h1:naHLuLoDiP4jHNo9R0sCBMtWGeIprob74mVsIT4qYEQ=
 github.com/stretchr/testify v1.11.1 h1:7s2iGBzp5EwR7/aIZr8ao5+dra3wiQyKjjFuvgVKu7U=
@@ -105,86 +93,76 @@ github.com/vishvananda/netlink v1.3.1 h1:3AEMt62VKqz90r0tmNhog0r/PpWKmrEShJU0wJW
 github.com/vishvananda/netlink v1.3.1/go.mod h1:ARtKouGSTGchR8aMwmkzC0qiNPrrWO5JS/XMVl45+b4=
 github.com/vishvananda/netns v0.0.5 h1:DfiHV+j8bA32MFM7bfEunvT8IAqQ/NzSJHtcmW5zdEY=
 github.com/vishvananda/netns v0.0.5/go.mod h1:SpkAiCQRtJ6TvvxPnOSyH3BMl6unz3xZlaprSwhNNJM=
-go.opentelemetry.io/auto/sdk v1.1.0 h1:cH53jehLUN6UFLY71z+NDOiNJqDdPRaXzTel0sJySYA=
-go.opentelemetry.io/auto/sdk v1.1.0/go.mod h1:3wSPjt5PWp2RhlCcmmOial7AvC4DQqZb7a7wCow3W8A=
-go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.63.0 h1:RbKq8BG0FI8OiXhBfcRtqqHcZcka+gU3cskNuf05R18=
-go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.63.0/go.mod h1:h06DGIukJOevXaj/xrNjhi/2098RZzcLTbc0jDAUbsg=
-go.opentelemetry.io/contrib/instrumentation/runtime v0.63.0 h1:PeBoRj6af6xMI7qCupwFvTbbnd49V7n5YpG6pg8iDYQ=
-go.opentelemetry.io/contrib/instrumentation/runtime v0.63.0/go.mod h1:ingqBCtMCe8I4vpz/UVzCW6sxoqgZB37nao91mLQ3Bw=
-go.opentelemetry.io/otel v1.38.0 h1:RkfdswUDRimDg0m2Az18RKOsnI8UDzppJAtj01/Ymk8=
-go.opentelemetry.io/otel v1.38.0/go.mod h1:zcmtmQ1+YmQM9wrNsTGV/q/uyusom3P8RxwExxkZhjM=
-go.opentelemetry.io/otel/exporters/otlp/otlpmetric/otlpmetricgrpc v1.38.0 h1:vl9obrcoWVKp/lwl8tRE33853I8Xru9HFbw/skNeLs8=
-go.opentelemetry.io/otel/exporters/otlp/otlpmetric/otlpmetricgrpc v1.38.0/go.mod h1:GAXRxmLJcVM3u22IjTg74zWBrRCKq8BnOqUVLodpcpw=
-go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.38.0 h1:GqRJVj7UmLjCVyVJ3ZFLdPRmhDUp2zFmQe3RHIOsw24=
-go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.38.0/go.mod h1:ri3aaHSmCTVYu2AWv44YMauwAQc0aqI9gHKIcSbI1pU=
-go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc v1.38.0 h1:lwI4Dc5leUqENgGuQImwLo4WnuXFPetmPpkLi2IrX54=
-go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc v1.38.0/go.mod h1:Kz/oCE7z5wuyhPxsXDuaPteSWqjSBD5YaSdbxZYGbGk=
+go.opentelemetry.io/auto/sdk v1.2.1 h1:jXsnJ4Lmnqd11kwkBV2LgLoFMZKizbCi5fNZ/ipaZ64=
+go.opentelemetry.io/auto/sdk v1.2.1/go.mod h1:KRTj+aOaElaLi+wW1kO/DZRXwkF4C5xPbEe3ZiIhN7Y=
+go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.66.0 h1:PnV4kVnw0zOmwwFkAzCN5O07fw1YOIQor120zrh0AVo=
+go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.66.0/go.mod h1:ofAwF4uinaf8SXdVzzbL4OsxJ3VfeEg3f/F6CeF49/Y=
+go.opentelemetry.io/contrib/instrumentation/runtime v0.66.0 h1:JruBNmrPELWjR+PU3fsQBFQRYtsMLQ/zPfbvwDz9I/w=
+go.opentelemetry.io/contrib/instrumentation/runtime v0.66.0/go.mod h1:vwNrfL6w1uAE3qX48KFii2Qoqf+NEDP5wNjus+RHz8Y=
+go.opentelemetry.io/otel v1.41.0 h1:YlEwVsGAlCvczDILpUXpIpPSL/VPugt7zHThEMLce1c=
+go.opentelemetry.io/otel v1.41.0/go.mod h1:Yt4UwgEKeT05QbLwbyHXEwhnjxNO6D8L5PQP51/46dE=
+go.opentelemetry.io/otel/exporters/otlp/otlpmetric/otlpmetricgrpc v1.41.0 h1:VO3BL6OZXRQ1yQc8W6EVfJzINeJ35BkiHx4MYfoQf44=
+go.opentelemetry.io/otel/exporters/otlp/otlpmetric/otlpmetricgrpc v1.41.0/go.mod h1:qRDnJ2nv3CQXMK2HUd9K9VtvedsPAce3S+/4LZHjX/s=
+go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.41.0 h1:ao6Oe+wSebTlQ1OEht7jlYTzQKE+pnx/iNywFvTbuuI=
+go.opentelemetry.io/otel/exporters/otlp/otlptrace v1.41.0/go.mod h1:u3T6vz0gh/NVzgDgiwkgLxpsSF6PaPmo2il0apGJbls=
+go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc v1.41.0 h1:mq/Qcf28TWz719lE3/hMB4KkyDuLJIvgJnFGcd0kEUI=
+go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc v1.41.0/go.mod h1:yk5LXEYhsL2htyDNJbEq7fWzNEigeEdV5xBF/Y+kAv0=
 go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp v1.38.0 h1:aTL7F04bJHUlztTsNGJ2l+6he8c+y/b//eR0jjjemT4=
 go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp v1.38.0/go.mod h1:kldtb7jDTeol0l3ewcmd8SDvx3EmIE7lyvqbasU3QC4=
-go.opentelemetry.io/otel/exporters/prometheus v0.60.0 h1:cGtQxGvZbnrWdC2GyjZi0PDKVSLWP/Jocix3QWfXtbo=
-go.opentelemetry.io/otel/exporters/prometheus v0.60.0/go.mod h1:hkd1EekxNo69PTV4OWFGZcKQiIqg0RfuWExcPKFvepk=
-go.opentelemetry.io/otel/metric v1.38.0 h1:Kl6lzIYGAh5M159u9NgiRkmoMKjvbsKtYRwgfrA6WpA=
-go.opentelemetry.io/otel/metric v1.38.0/go.mod h1:kB5n/QoRM8YwmUahxvI3bO34eVtQf2i4utNVLr9gEmI=
-go.opentelemetry.io/otel/sdk v1.38.0 h1:l48sr5YbNf2hpCUj/FoGhW9yDkl+Ma+LrVl8qaM5b+E=
-go.opentelemetry.io/otel/sdk v1.38.0/go.mod h1:ghmNdGlVemJI3+ZB5iDEuk4bWA3GkTpW+DOoZMYBVVg=
-go.opentelemetry.io/otel/sdk/metric v1.38.0 h1:aSH66iL0aZqo//xXzQLYozmWrXxyFkBJ6qT5wthqPoM=
-go.opentelemetry.io/otel/sdk/metric v1.38.0/go.mod h1:dg9PBnW9XdQ1Hd6ZnRz689CbtrUp0wMMs9iPcgT9EZA=
-go.opentelemetry.io/otel/trace v1.38.0 h1:Fxk5bKrDZJUH+AMyyIXGcFAPah0oRcT+LuNtJrmcNLE=
-go.opentelemetry.io/otel/trace v1.38.0/go.mod h1:j1P9ivuFsTceSWe1oY+EeW3sc+Pp42sO++GHkg4wwhs=
-go.opentelemetry.io/proto/otlp v1.7.1 h1:gTOMpGDb0WTBOP8JaO72iL3auEZhVmAQg4ipjOVAtj4=
-go.opentelemetry.io/proto/otlp v1.7.1/go.mod h1:b2rVh6rfI/s2pHWNlB7ILJcRALpcNDzKhACevjI+ZnE=
+go.opentelemetry.io/otel/exporters/prometheus v0.63.0 h1:OLo1FNb0pBZykLqbKRZolKtGZd0Waqlr240YdMEnhhg=
+go.opentelemetry.io/otel/exporters/prometheus v0.63.0/go.mod h1:8yeQAdhrK5xsWuFehO13Dk/Xb9FuhZoVpJfpoNCfJnw=
+go.opentelemetry.io/otel/metric v1.41.0 h1:rFnDcs4gRzBcsO9tS8LCpgR0dxg4aaxWlJxCno7JlTQ=
+go.opentelemetry.io/otel/metric v1.41.0/go.mod h1:xPvCwd9pU0VN8tPZYzDZV/BMj9CM9vs00GuBjeKhJps=
+go.opentelemetry.io/otel/sdk v1.41.0 h1:YPIEXKmiAwkGl3Gu1huk1aYWwtpRLeskpV+wPisxBp8=
+go.opentelemetry.io/otel/sdk v1.41.0/go.mod h1:ahFdU0G5y8IxglBf0QBJXgSe7agzjE4GiTJ6HT9ud90=
+go.opentelemetry.io/otel/sdk/metric v1.41.0 h1:siZQIYBAUd1rlIWQT2uCxWJxcCO7q3TriaMlf08rXw8=
+go.opentelemetry.io/otel/sdk/metric v1.41.0/go.mod h1:HNBuSvT7ROaGtGI50ArdRLUnvRTRGniSUZbxiWxSO8Y=
+go.opentelemetry.io/otel/trace v1.41.0 h1:Vbk2co6bhj8L59ZJ6/xFTskY+tGAbOnCtQGVVa9TIN0=
+go.opentelemetry.io/otel/trace v1.41.0/go.mod h1:U1NU4ULCoxeDKc09yCWdWe+3QoyweJcISEVa1RBzOis=
+go.opentelemetry.io/proto/otlp v1.9.0 h1:l706jCMITVouPOqEnii2fIAuO3IVGBRPV5ICjceRb/A=
+go.opentelemetry.io/proto/otlp v1.9.0/go.mod h1:xE+Cx5E/eEHw+ISFkwPLwCZefwVjY+pqKg1qcK03+/4=
 go.uber.org/goleak v1.3.0 h1:2K3zAYmnTNqV73imy9J1T3WC+gmCePx2hEGkimedGto=
 go.uber.org/goleak v1.3.0/go.mod h1:CoHD4mav9JJNrW/WLlf7HGZPjdw8EucARQHekz1X6bE=
-go.yaml.in/yaml/v2 v2.4.2 h1:DzmwEr2rDGHl7lsFgAHxmNz/1NlQ7xLIrlN2h5d1eGI=
-go.yaml.in/yaml/v2 v2.4.2/go.mod h1:081UH+NErpNdqlCXm3TtEran0rJZGxAYx9hb/ELlsPU=
-golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
-golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
-golang.org/x/crypto v0.44.0 h1:A97SsFvM3AIwEEmTBiaxPPTYpDC47w720rdiiUvgoAU=
-golang.org/x/crypto v0.44.0/go.mod h1:013i+Nw79BMiQiMsOPcVCB5ZIJbYkerPrGnOa00tvmc=
+go.yaml.in/yaml/v2 v2.4.3 h1:6gvOSjQoTB3vt1l+CU+tSyi/HOjfOjRLJ4YwYZGwRO0=
+go.yaml.in/yaml/v2 v2.4.3/go.mod h1:zSxWcmIDjOzPXpjlTTbAsKokqkDNAVtZO0WOMiT90s8=
+golang.org/x/crypto v0.48.0 h1:/VRzVqiRSggnhY7gNRxPauEQ5Drw9haKdM0jqfcCFts=
+golang.org/x/crypto v0.48.0/go.mod h1:r0kV5h3qnFPlQnBSrULhlsRfryS2pmewsg+XfMgkVos=
 golang.org/x/exp v0.0.0-20251113190631-e25ba8c21ef6 h1:zfMcR1Cs4KNuomFFgGefv5N0czO2XZpUbxGUy8i8ug0=
 golang.org/x/exp v0.0.0-20251113190631-e25ba8c21ef6/go.mod h1:46edojNIoXTNOhySWIWdix628clX9ODXwPsQuG6hsK0=
-golang.org/x/lint v0.0.0-20200302205851-738671d3881b/go.mod h1:3xt1FjdF8hUf6vQPIChWIBhFzV8gjjsPE/fR3IyQdNY=
-golang.org/x/mod v0.1.1-0.20191105210325-c90efee705ee/go.mod h1:QqPTAvyqsEbceGzBzNggFXnrqF1CaUcvgkdR5Ot7KZg=
-golang.org/x/mod v0.30.0 h1:fDEXFVZ/fmCKProc/yAXXUijritrDzahmwwefnjoPFk=
-golang.org/x/mod v0.30.0/go.mod h1:lAsf5O2EvJeSFMiBxXDki7sCgAxEUcZHXoXMKT4GJKc=
-golang.org/x/net v0.0.0-20190404232315-eb5bcb51f2a3/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg=
-golang.org/x/net v0.0.0-20190620200207-3b0461eec859/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
-golang.org/x/net v0.47.0 h1:Mx+4dIFzqraBXUugkia1OOvlD6LemFo1ALMHjrXDOhY=
-golang.org/x/net v0.47.0/go.mod h1:/jNxtkgq5yWUGYkaZGqo27cfGZ1c5Nen03aYrrKpVRU=
-golang.org/x/sync v0.0.0-20190423024810-112230192c58/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
-golang.org/x/sync v0.18.0 h1:kr88TuHDroi+UVf+0hZnirlk8o8T+4MrK6mr60WkH/I=
-golang.org/x/sync v0.18.0/go.mod h1:9KTHXmSnoGruLpwFjVSX0lNNA75CykiMECbovNTZqGI=
-golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
-golang.org/x/sys v0.0.0-20190412213103-97732733099d/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/mod v0.32.0 h1:9F4d3PHLljb6x//jOyokMv3eX+YDeepZSEo3mFJy93c=
+golang.org/x/mod v0.32.0/go.mod h1:SgipZ/3h2Ci89DlEtEXWUk/HteuRin+HHhN+WbNhguU=
+golang.org/x/net v0.51.0 h1:94R/GTO7mt3/4wIKpcR5gkGmRLOuE/2hNGeWq/GBIFo=
+golang.org/x/net v0.51.0/go.mod h1:aamm+2QF5ogm02fjy5Bb7CQ0WMt1/WVM7FtyaTLlA9Y=
+golang.org/x/sync v0.19.0 h1:vV+1eWNmZ5geRlYjzm2adRgW2/mcpevXNg50YZtPCE4=
+golang.org/x/sync v0.19.0/go.mod h1:9KTHXmSnoGruLpwFjVSX0lNNA75CykiMECbovNTZqGI=
 golang.org/x/sys v0.2.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.10.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.38.0 h1:3yZWxaJjBmCWXqhN1qh02AkOnCQ1poK6oF+a7xWL6Gc=
-golang.org/x/sys v0.38.0/go.mod h1:OgkHotnGiDImocRcuBABYBEXf8A9a87e/uXjp9XT3ks=
-golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
-golang.org/x/text v0.31.0 h1:aC8ghyu4JhP8VojJ2lEHBnochRno1sgL6nEi9WGFGMM=
-golang.org/x/text v0.31.0/go.mod h1:tKRAlv61yKIjGGHX/4tP1LTbc13YSec1pxVEWXzfoeM=
+golang.org/x/sys v0.41.0 h1:Ivj+2Cp/ylzLiEU89QhWblYnOE9zerudt9Ftecq2C6k=
+golang.org/x/sys v0.41.0/go.mod h1:OgkHotnGiDImocRcuBABYBEXf8A9a87e/uXjp9XT3ks=
+golang.org/x/text v0.34.0 h1:oL/Qq0Kdaqxa1KbNeMKwQq0reLCCaFtqu2eNuSeNHbk=
+golang.org/x/text v0.34.0/go.mod h1:homfLqTYRFyVYemLBFl5GgL/DWEiH5wcsQ5gSh1yziA=
 golang.org/x/time v0.12.0 h1:ScB/8o8olJvc+CQPWrK3fPZNfh7qgwCrY0zJmoEQLSE=
 golang.org/x/time v0.12.0/go.mod h1:CDIdPxbZBQxdj6cxyCIdrNogrJKMJ7pr37NYpMcMDSg=
-golang.org/x/tools v0.0.0-20200130002326-2f3ba24bd6e7/go.mod h1:TB2adYChydJhpapKDTa4BR/hXlZSLoq2Wpct/0txZ28=
-golang.org/x/tools v0.39.0 h1:ik4ho21kwuQln40uelmciQPp9SipgNDdrafrYA4TmQQ=
-golang.org/x/tools v0.39.0/go.mod h1:JnefbkDPyD8UU2kI5fuf8ZX4/yUeh9W877ZeBONxUqQ=
-golang.org/x/xerrors v0.0.0-20191011141410-1b5146add898/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
+golang.org/x/tools v0.41.0 h1:a9b8iMweWG+S0OBnlU36rzLp20z1Rp10w+IY2czHTQc=
+golang.org/x/tools v0.41.0/go.mod h1:XSY6eDqxVNiYgezAVqqCeihT4j1U2CCsqvH3WhQpnlg=
 golang.zx2c4.com/wintun v0.0.0-20230126152724-0fa3db229ce2 h1:B82qJJgjvYKsXS9jeunTOisW56dUokqW/FOteYJJ/yg=
 golang.zx2c4.com/wintun v0.0.0-20230126152724-0fa3db229ce2/go.mod h1:deeaetjYA+DHMHg+sMSMI58GrEteJUUzzw7en6TJQcI=
 golang.zx2c4.com/wireguard v0.0.0-20250521234502-f333402bd9cb h1:whnFRlWMcXI9d+ZbWg+4sHnLp52d5yiIPUxMBSt4X9A=
 golang.zx2c4.com/wireguard v0.0.0-20250521234502-f333402bd9cb/go.mod h1:rpwXGsirqLqN2L0JDJQlwOboGHmptD5ZD6T2VmcqhTw=
 golang.zx2c4.com/wireguard/wgctrl v0.0.0-20241231184526-a9ab2273dd10 h1:3GDAcqdIg1ozBNLgPy4SLT84nfcBjr6rhGtXYtrkWLU=
 golang.zx2c4.com/wireguard/wgctrl v0.0.0-20241231184526-a9ab2273dd10/go.mod h1:T97yPqesLiNrOYxkwmhMI0ZIlJDm+p0PMR8eRVeR5tQ=
+golang.zx2c4.com/wireguard/windows v0.5.3 h1:On6j2Rpn3OEMXqBq00QEDC7bWSZrPIHKIus8eIuExIE=
+golang.zx2c4.com/wireguard/windows v0.5.3/go.mod h1:9TEe8TJmtwyQebdFwAkEWOPr3prrtqm+REGFifP60hI=
 gonum.org/v1/gonum v0.16.0 h1:5+ul4Swaf3ESvrOnidPp4GZbzf0mxVQpDCYUQE7OJfk=
 gonum.org/v1/gonum v0.16.0/go.mod h1:fef3am4MQ93R2HHpKnLk4/Tbh/s0+wqD5nfa6Pnwy4E=
-google.golang.org/genproto/googleapis/api v0.0.0-20250825161204-c5933d9347a5 h1:BIRfGDEjiHRrk0QKZe3Xv2ieMhtgRGeLcZQ0mIVn4EY=
-google.golang.org/genproto/googleapis/api v0.0.0-20250825161204-c5933d9347a5/go.mod h1:j3QtIyytwqGr1JUDtYXwtMXWPKsEa5LtzIFN1Wn5WvE=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20250825161204-c5933d9347a5 h1:eaY8u2EuxbRv7c3NiGK0/NedzVsCcV6hDuU5qPX5EGE=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20250825161204-c5933d9347a5/go.mod h1:M4/wBTSeyLxupu3W3tJtOgB14jILAS/XWPSSa3TAlJc=
-google.golang.org/grpc v1.76.0 h1:UnVkv1+uMLYXoIz6o7chp59WfQUYA2ex/BXQ9rHZu7A=
-google.golang.org/grpc v1.76.0/go.mod h1:Ju12QI8M6iQJtbcsV+awF5a4hfJMLi4X0JLo94ULZ6c=
-google.golang.org/protobuf v1.36.8 h1:xHScyCOEuuwZEc6UtSOvPbAT4zRh0xcNRYekJwfqyMc=
-google.golang.org/protobuf v1.36.8/go.mod h1:fuxRtAxBytpl4zzqUh6/eyUujkJdNiuEkXntxiD/uRU=
+google.golang.org/genproto/googleapis/api v0.0.0-20260209200024-4cfbd4190f57 h1:JLQynH/LBHfCTSbDWl+py8C+Rg/k1OVH3xfcaiANuF0=
+google.golang.org/genproto/googleapis/api v0.0.0-20260209200024-4cfbd4190f57/go.mod h1:kSJwQxqmFXeo79zOmbrALdflXQeAYcUbgS7PbpMknCY=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20260209200024-4cfbd4190f57 h1:mWPCjDEyshlQYzBpMNHaEof6UX1PmHcaUODUywQ0uac=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20260209200024-4cfbd4190f57/go.mod h1:j9x/tPzZkyxcgEFkiKEEGxfvyumM01BEtsW8xzOahRQ=
+google.golang.org/grpc v1.79.1 h1:zGhSi45ODB9/p3VAawt9a+O/MULLl9dpizzNNpq7flY=
+google.golang.org/grpc v1.79.1/go.mod h1:KmT0Kjez+0dde/v2j9vzwoAScgEPx/Bw1CYChhHLrHQ=
+google.golang.org/protobuf v1.36.11 h1:fV6ZwhNocDyBLK0dj+fg8ektcVegBBuEolpbTQyBNVE=
+google.golang.org/protobuf v1.36.11/go.mod h1:HTf+CrKn2C3g5S8VImy6tdcUvCska2kB7j23XfzDpco=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c/go.mod h1:JHkPIbrfpd72SG/EVd6muEfDQjcINNoR0C8j2r3qZ4Q=
@@ -194,5 +172,5 @@ gotest.tools/v3 v3.4.0 h1:ZazjZUfuVeZGLAmlKKuyv3IKP5orXcwtOwDQH6YVr6o=
 gotest.tools/v3 v3.4.0/go.mod h1:CtbdzLSsqVhDgMtKsx03ird5YTGB3ar27v0u/yKBW5g=
 gvisor.dev/gvisor v0.0.0-20250503011706-39ed1f5ac29c h1:m/r7OM+Y2Ty1sgBQ7Qb27VgIMBW8ZZhT4gLnUyDIhzI=
 gvisor.dev/gvisor v0.0.0-20250503011706-39ed1f5ac29c/go.mod h1:3r5CMtNQMKIvBlrmM9xWUNamjKBYPOWyXOjmg5Kts3g=
-software.sslmate.com/src/go-pkcs12 v0.6.0 h1:f3sQittAeF+pao32Vb+mkli+ZyT+VwKaD014qFGq6oU=
-software.sslmate.com/src/go-pkcs12 v0.6.0/go.mod h1:Qiz0EyvDRJjjxGyUQa2cCNZn/wMyzrRJ/qcDXOQazLI=
+software.sslmate.com/src/go-pkcs12 v0.7.0 h1:Db8W44cB54TWD7stUFFSWxdfpdn6fZVcDl0w3R4RVM0=
+software.sslmate.com/src/go-pkcs12 v0.7.0/go.mod h1:Qiz0EyvDRJjjxGyUQa2cCNZn/wMyzrRJ/qcDXOQazLI=

healthcheck/healthcheck.go

@@ -48,6 +48,7 @@ type Config struct {
 	Headers           map[string]string `json:"hcHeaders"`
 	Method            string            `json:"hcMethod"`
 	Status            int               `json:"hcStatus"` // HTTP status code
+	TLSServerName     string            `json:"hcTlsServerName"`
 }
 
 // Target represents a health check target with its current status
@@ -57,9 +58,10 @@ type Target struct {
 	LastCheck  time.Time `json:"lastCheck"`
 	LastError  string    `json:"lastError,omitempty"`
 	CheckCount int       `json:"checkCount"`
-	ticker     *time.Ticker
+	timer      *time.Timer
 	ctx        context.Context
 	cancel     context.CancelFunc
+	client     *http.Client
 }
 
 // StatusChangeCallback is called when any target's status changes
@@ -70,7 +72,6 @@ type Monitor struct {
 	targets     map[int]*Target
 	mutex       sync.RWMutex
 	callback    StatusChangeCallback
-	client      *http.Client
 	enforceCert bool
 }
 
@@ -78,21 +79,10 @@ type Monitor struct {
 func NewMonitor(callback StatusChangeCallback, enforceCert bool) *Monitor {
 	logger.Debug("Creating new health check monitor with certificate enforcement: %t", enforceCert)
 
-	// Configure TLS settings based on certificate enforcement
-	transport := &http.Transport{
-		TLSClientConfig: &tls.Config{
-			InsecureSkipVerify: !enforceCert,
-		},
-	}
-
 	return &Monitor{
 		targets:     make(map[int]*Target),
 		callback:    callback,
 		enforceCert: enforceCert,
-		client: &http.Client{
-			Timeout:   30 * time.Second,
-			Transport: transport,
-		},
 	}
 }
 
@@ -196,6 +186,16 @@ func (m *Monitor) addTargetUnsafe(config Config) error {
 		Status: StatusUnknown,
 		ctx:    ctx,
 		cancel: cancel,
+		client: &http.Client{
+			Transport: &http.Transport{
+				TLSClientConfig: &tls.Config{
+					// Configure TLS settings based on certificate enforcement
+					InsecureSkipVerify: !m.enforceCert,
+					// Use SNI TLS header if present
+					ServerName: config.TLSServerName,
+				},
+			},
+		},
 	}
 
 	m.targets[config.ID] = target
@@ -315,26 +315,26 @@ func (m *Monitor) monitorTarget(target *Target) {
 		go m.callback(m.GetTargets())
 	}
 
-	// Set up ticker based on current status
+	// Set up timer based on current status
 	interval := time.Duration(target.Config.Interval) * time.Second
 	if target.Status == StatusUnhealthy {
 		interval = time.Duration(target.Config.UnhealthyInterval) * time.Second
 	}
 
 	logger.Debug("Target %d: initial check interval set to %v", target.Config.ID, interval)
-	target.ticker = time.NewTicker(interval)
-	defer target.ticker.Stop()
+	target.timer = time.NewTimer(interval)
+	defer target.timer.Stop()
 
 	for {
 		select {
 		case <-target.ctx.Done():
 			logger.Info("Stopping health check monitoring for target %d", target.Config.ID)
 			return
-		case <-target.ticker.C:
+		case <-target.timer.C:
 			oldStatus := target.Status
 			m.performHealthCheck(target)
 
-			// Update ticker interval if status changed
+			// Update timer interval if status changed
 			newInterval := time.Duration(target.Config.Interval) * time.Second
 			if target.Status == StatusUnhealthy {
 				newInterval = time.Duration(target.Config.UnhealthyInterval) * time.Second
@@ -343,11 +343,12 @@ func (m *Monitor) monitorTarget(target *Target) {
 			if newInterval != interval {
 				logger.Debug("Target %d: updating check interval from %v to %v due to status change",
 					target.Config.ID, interval, newInterval)
-				target.ticker.Stop()
-				target.ticker = time.NewTicker(newInterval)
 				interval = newInterval
 			}
 
+			// Reset timer for next check with current interval
+			target.timer.Reset(interval)
+
 			// Notify callback if status changed
 			if oldStatus != target.Status && m.callback != nil {
 				logger.Info("Target %d status changed: %s -> %s",
@@ -398,11 +399,16 @@ func (m *Monitor) performHealthCheck(target *Target) {
 
 	// Add headers
 	for key, value := range target.Config.Headers {
-		req.Header.Set(key, value)
+		// Handle Host header specially - it must be set on req.Host, not in headers
+		if strings.EqualFold(key, "Host") {
+			req.Host = value
+		} else {
+			req.Header.Set(key, value)
+		}
 	}
 
 	// Perform request
-	resp, err := m.client.Do(req)
+	resp, err := target.client.Do(req)
 	if err != nil {
 		target.Status = StatusUnhealthy
 		target.LastError = fmt.Sprintf("request failed: %v", err)

holepunch/holepunch.go

@@ -4,6 +4,7 @@ import (
 	"encoding/json"
 	"fmt"
 	"net"
+	"strconv"
 	"sync"
 	"time"
 
@@ -19,7 +20,9 @@ import (
 // ExitNode represents a WireGuard exit node for hole punching
 type ExitNode struct {
 	Endpoint  string `json:"endpoint"`
+	RelayPort uint16 `json:"relayPort"`
 	PublicKey string `json:"publicKey"`
+	SiteIds   []int  `json:"siteIds,omitempty"`
 }
 
 // Manager handles UDP hole punching operations
@@ -35,21 +38,29 @@ type Manager struct {
 	exitNodes  map[string]ExitNode // key is endpoint
 	updateChan chan struct{}       // signals the goroutine to refresh exit nodes
 
-	sendHolepunchInterval time.Duration
+	sendHolepunchInterval    time.Duration
+	sendHolepunchIntervalMin time.Duration
+	sendHolepunchIntervalMax time.Duration
+	defaultIntervalMin       time.Duration
+	defaultIntervalMax       time.Duration
 }
 
-const sendHolepunchIntervalMax = 60 * time.Second
-const sendHolepunchIntervalMin = 1 * time.Second
+const defaultSendHolepunchIntervalMax = 60 * time.Second
+const defaultSendHolepunchIntervalMin = 1 * time.Second
 
 // NewManager creates a new hole punch manager
 func NewManager(sharedBind *bind.SharedBind, ID string, clientType string, publicKey string) *Manager {
 	return &Manager{
-		sharedBind:            sharedBind,
-		ID:                    ID,
-		clientType:            clientType,
-		publicKey:             publicKey,
-		exitNodes:             make(map[string]ExitNode),
-		sendHolepunchInterval: sendHolepunchIntervalMin,
+		sharedBind:               sharedBind,
+		ID:                       ID,
+		clientType:               clientType,
+		publicKey:                publicKey,
+		exitNodes:                make(map[string]ExitNode),
+		sendHolepunchInterval:    defaultSendHolepunchIntervalMin,
+		sendHolepunchIntervalMin: defaultSendHolepunchIntervalMin,
+		sendHolepunchIntervalMax: defaultSendHolepunchIntervalMax,
+		defaultIntervalMin:       defaultSendHolepunchIntervalMin,
+		defaultIntervalMax:       defaultSendHolepunchIntervalMax,
 	}
 }
 
@@ -140,6 +151,51 @@ func (m *Manager) RemoveExitNode(endpoint string) bool {
 	return true
 }
 
+/*
+RemoveExitNodesByPeer removes the peer ID from the SiteIds list in each exit node.
+If the SiteIds list becomes empty after removal, the exit node is removed entirely.
+Returns the number of exit nodes removed.
+*/
+func (m *Manager) RemoveExitNodesByPeer(peerID int) int {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	removed := 0
+	for endpoint, node := range m.exitNodes {
+		// Remove peerID from SiteIds if present
+		newSiteIds := make([]int, 0, len(node.SiteIds))
+		for _, id := range node.SiteIds {
+			if id != peerID {
+				newSiteIds = append(newSiteIds, id)
+			}
+		}
+		if len(newSiteIds) != len(node.SiteIds) {
+			node.SiteIds = newSiteIds
+			if len(node.SiteIds) == 0 {
+				delete(m.exitNodes, endpoint)
+				logger.Info("Removed exit node %s as no more site IDs remain after removing peer %d", endpoint, peerID)
+				removed++
+			} else {
+				m.exitNodes[endpoint] = node
+				logger.Info("Removed peer %d from exit node %s site IDs", peerID, endpoint)
+			}
+		}
+	}
+
+	if removed > 0 {
+		// Signal the goroutine to refresh if running
+		if m.running && m.updateChan != nil {
+			select {
+			case m.updateChan <- struct{}{}:
+			default:
+				// Channel full or closed, skip
+			}
+		}
+	}
+
+	return removed
+}
+
 // GetExitNodes returns a copy of the current exit nodes
 func (m *Manager) GetExitNodes() []ExitNode {
 	m.mu.Lock()
@@ -152,17 +208,46 @@ func (m *Manager) GetExitNodes() []ExitNode {
 	return nodes
 }
 
-// ResetInterval resets the hole punch interval back to the minimum value,
-// allowing it to climb back up through exponential backoff.
-// This is useful when network conditions change or connectivity is restored.
-func (m *Manager) ResetInterval() {
+// SetServerHolepunchInterval sets custom min and max intervals for hole punching.
+// This is useful for low power mode where longer intervals are desired.
+func (m *Manager) SetServerHolepunchInterval(min, max time.Duration) {
 	m.mu.Lock()
 	defer m.mu.Unlock()
 
-	if m.sendHolepunchInterval != sendHolepunchIntervalMin {
-		m.sendHolepunchInterval = sendHolepunchIntervalMin
-		logger.Info("Reset hole punch interval to minimum (%v)", sendHolepunchIntervalMin)
+	m.sendHolepunchIntervalMin = min
+	m.sendHolepunchIntervalMax = max
+	m.sendHolepunchInterval = min
+
+	logger.Info("Set hole punch intervals: min=%v, max=%v", min, max)
+
+	// Signal the goroutine to apply the new interval if running
+	if m.running && m.updateChan != nil {
+		select {
+		case m.updateChan <- struct{}{}:
+		default:
+			// Channel full or closed, skip
+		}
 	}
+}
+
+// GetInterval returns the current min and max intervals
+func (m *Manager) GetServerHolepunchInterval() (min, max time.Duration) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	return m.sendHolepunchIntervalMin, m.sendHolepunchIntervalMax
+}
+
+// ResetServerHolepunchInterval resets the hole punch interval back to the default values.
+// This restores normal operation after low power mode or other custom settings.
+func (m *Manager) ResetServerHolepunchInterval() {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	m.sendHolepunchIntervalMin = m.defaultIntervalMin
+	m.sendHolepunchIntervalMax = m.defaultIntervalMax
+	m.sendHolepunchInterval = m.defaultIntervalMin
+
+	logger.Info("Reset hole punch intervals to defaults: min=%v, max=%v", m.defaultIntervalMin, m.defaultIntervalMax)
 
 	// Signal the goroutine to apply the new interval if running
 	if m.running && m.updateChan != nil {
@@ -202,7 +287,7 @@ func (m *Manager) TriggerHolePunch() error {
 			continue
 		}
 
-		serverAddr := net.JoinHostPort(host, "21820")
+		serverAddr := net.JoinHostPort(host, strconv.Itoa(int(exitNode.RelayPort)))
 		remoteAddr, err := net.ResolveUDPAddr("udp", serverAddr)
 		if err != nil {
 			logger.Error("Failed to resolve UDP address %s: %v", serverAddr, err)
@@ -236,12 +321,6 @@ func (m *Manager) StartMultipleExitNodes(exitNodes []ExitNode) error {
 		return fmt.Errorf("hole punch already running")
 	}
 
-	if len(exitNodes) == 0 {
-		m.mu.Unlock()
-		logger.Warn("No exit nodes provided for hole punching")
-		return fmt.Errorf("no exit nodes provided")
-	}
-
 	// Populate exit nodes map
 	m.exitNodes = make(map[string]ExitNode)
 	for _, node := range exitNodes {
@@ -253,7 +332,7 @@ func (m *Manager) StartMultipleExitNodes(exitNodes []ExitNode) error {
 	m.updateChan = make(chan struct{}, 1)
 	m.mu.Unlock()
 
-	logger.Info("Starting UDP hole punch to %d exit nodes with shared bind", len(exitNodes))
+	logger.Debug("Starting UDP hole punch to %d exit nodes with shared bind", len(exitNodes))
 
 	go m.runMultipleExitNodes()
 
@@ -270,18 +349,17 @@ func (m *Manager) Start() error {
 		return fmt.Errorf("hole punch already running")
 	}
 
-	if len(m.exitNodes) == 0 {
-		m.mu.Unlock()
-		logger.Warn("No exit nodes configured for hole punching")
-		return fmt.Errorf("no exit nodes configured")
-	}
-
 	m.running = true
 	m.stopChan = make(chan struct{})
 	m.updateChan = make(chan struct{}, 1)
+	nodeCount := len(m.exitNodes)
 	m.mu.Unlock()
 
-	logger.Info("Starting UDP hole punch with %d exit nodes", len(m.exitNodes))
+	if nodeCount == 0 {
+		logger.Info("Starting UDP hole punch manager (waiting for exit nodes to be added)")
+	} else {
+		logger.Info("Starting UDP hole punch with %d exit nodes", nodeCount)
+	}
 
 	go m.runMultipleExitNodes()
 
@@ -320,7 +398,7 @@ func (m *Manager) runMultipleExitNodes() {
 				continue
 			}
 
-			serverAddr := net.JoinHostPort(host, "21820")
+			serverAddr := net.JoinHostPort(host, strconv.Itoa(int(exitNode.RelayPort)))
 			remoteAddr, err := net.ResolveUDPAddr("udp", serverAddr)
 			if err != nil {
 				logger.Error("Failed to resolve UDP address %s: %v", serverAddr, err)
@@ -332,7 +410,7 @@ func (m *Manager) runMultipleExitNodes() {
 				publicKey:    exitNode.PublicKey,
 				endpointName: exitNode.Endpoint,
 			})
-			logger.Info("Resolved exit node: %s -> %s", exitNode.Endpoint, remoteAddr.String())
+			logger.Debug("Resolved exit node: %s -> %s", exitNode.Endpoint, remoteAddr.String())
 		}
 		return resolvedNodes
 	}
@@ -340,20 +418,19 @@ func (m *Manager) runMultipleExitNodes() {
 	resolvedNodes := resolveNodes()
 
 	if len(resolvedNodes) == 0 {
-		logger.Error("No exit nodes could be resolved")
-		return
-	}
-
-	// Send initial hole punch to all exit nodes
-	for _, node := range resolvedNodes {
-		if err := m.sendHolePunch(node.remoteAddr, node.publicKey); err != nil {
-			logger.Warn("Failed to send initial hole punch to %s: %v", node.endpointName, err)
+		logger.Info("No exit nodes available yet, waiting for nodes to be added")
+	} else {
+		// Send initial hole punch to all exit nodes
+		for _, node := range resolvedNodes {
+			if err := m.sendHolePunch(node.remoteAddr, node.publicKey); err != nil {
+				logger.Warn("Failed to send initial hole punch to %s: %v", node.endpointName, err)
+			}
 		}
 	}
 
 	// Start with minimum interval
 	m.mu.Lock()
-	m.sendHolepunchInterval = sendHolepunchIntervalMin
+	m.sendHolepunchInterval = m.sendHolepunchIntervalMin
 	m.mu.Unlock()
 
 	ticker := time.NewTicker(m.sendHolepunchInterval)
@@ -370,10 +447,12 @@ func (m *Manager) runMultipleExitNodes() {
 			resolvedNodes = resolveNodes()
 			if len(resolvedNodes) == 0 {
 				logger.Warn("No exit nodes available after refresh")
+			} else {
+				logger.Info("Updated resolved nodes count: %d", len(resolvedNodes))
 			}
 			// Reset interval to minimum on update
 			m.mu.Lock()
-			m.sendHolepunchInterval = sendHolepunchIntervalMin
+			m.sendHolepunchInterval = m.sendHolepunchIntervalMin
 			m.mu.Unlock()
 			ticker.Reset(m.sendHolepunchInterval)
 			// Send immediate hole punch to newly resolved nodes
@@ -383,24 +462,26 @@ func (m *Manager) runMultipleExitNodes() {
 				}
 			}
 		case <-ticker.C:
-			// Send hole punch to all exit nodes
-			for _, node := range resolvedNodes {
-				if err := m.sendHolePunch(node.remoteAddr, node.publicKey); err != nil {
-					logger.Debug("Failed to send hole punch to %s: %v", node.endpointName, err)
+			// Send hole punch to all exit nodes (if any are available)
+			if len(resolvedNodes) > 0 {
+				for _, node := range resolvedNodes {
+					if err := m.sendHolePunch(node.remoteAddr, node.publicKey); err != nil {
+						logger.Debug("Failed to send hole punch to %s: %v", node.endpointName, err)
+					}
 				}
+				// Exponential backoff: double the interval up to max
+				m.mu.Lock()
+				newInterval := m.sendHolepunchInterval * 2
+				if newInterval > m.sendHolepunchIntervalMax {
+					newInterval = m.sendHolepunchIntervalMax
+				}
+				if newInterval != m.sendHolepunchInterval {
+					m.sendHolepunchInterval = newInterval
+					ticker.Reset(m.sendHolepunchInterval)
+					logger.Debug("Increased hole punch interval to %v", m.sendHolepunchInterval)
+				}
+				m.mu.Unlock()
 			}
-			// Exponential backoff: double the interval up to max
-			m.mu.Lock()
-			newInterval := m.sendHolepunchInterval * 2
-			if newInterval > sendHolepunchIntervalMax {
-				newInterval = sendHolepunchIntervalMax
-			}
-			if newInterval != m.sendHolepunchInterval {
-				m.sendHolepunchInterval = newInterval
-				ticker.Reset(m.sendHolepunchInterval)
-				logger.Debug("Increased hole punch interval to %v", m.sendHolepunchInterval)
-			}
-			m.mu.Unlock()
 		}
 	}
 }

holepunch/tester.go

@@ -41,6 +41,12 @@ func DefaultTestOptions() TestConnectionOptions {
 	}
 }
 
+// cachedAddr holds a cached resolved UDP address
+type cachedAddr struct {
+	addr      *net.UDPAddr
+	resolvedAt time.Time
+}
+
 // HolepunchTester monitors holepunch connectivity using magic packets
 type HolepunchTester struct {
 	sharedBind *bind.SharedBind
@@ -53,6 +59,11 @@ type HolepunchTester struct {
 
 	// Callback when connection status changes
 	callback HolepunchStatusCallback
+
+	// Address cache to avoid repeated DNS/UDP resolution
+	addrCache    map[string]*cachedAddr
+	addrCacheMu  sync.RWMutex
+	addrCacheTTL time.Duration // How long cached addresses are valid
 }
 
 // HolepunchStatus represents the status of a holepunch connection
@@ -75,7 +86,9 @@ type pendingRequest struct {
 // NewHolepunchTester creates a new holepunch tester using the given SharedBind
 func NewHolepunchTester(sharedBind *bind.SharedBind) *HolepunchTester {
 	return &HolepunchTester{
-		sharedBind: sharedBind,
+		sharedBind:   sharedBind,
+		addrCache:    make(map[string]*cachedAddr),
+		addrCacheTTL: 5 * time.Minute, // Cache addresses for 5 minutes
 	}
 }
 
@@ -135,12 +148,70 @@ func (t *HolepunchTester) Stop() {
 		return true
 	})
 
+	// Clear address cache
+	t.addrCacheMu.Lock()
+	t.addrCache = make(map[string]*cachedAddr)
+	t.addrCacheMu.Unlock()
+
 	logger.Debug("HolepunchTester stopped")
 }
 
+// resolveEndpoint resolves an endpoint to a UDP address, using cache when possible
+func (t *HolepunchTester) resolveEndpoint(endpoint string) (*net.UDPAddr, error) {
+	// Check cache first
+	t.addrCacheMu.RLock()
+	cached, ok := t.addrCache[endpoint]
+	ttl := t.addrCacheTTL
+	t.addrCacheMu.RUnlock()
+
+	if ok && time.Since(cached.resolvedAt) < ttl {
+		return cached.addr, nil
+	}
+
+	// Resolve the endpoint
+	host, err := util.ResolveDomain(endpoint)
+	if err != nil {
+		host = endpoint
+	}
+
+	_, _, err = net.SplitHostPort(host)
+	if err != nil {
+		host = net.JoinHostPort(host, "21820")
+	}
+
+	remoteAddr, err := net.ResolveUDPAddr("udp", host)
+	if err != nil {
+		return nil, fmt.Errorf("failed to resolve UDP address %s: %w", host, err)
+	}
+
+	// Cache the result
+	t.addrCacheMu.Lock()
+	t.addrCache[endpoint] = &cachedAddr{
+		addr:       remoteAddr,
+		resolvedAt: time.Now(),
+	}
+	t.addrCacheMu.Unlock()
+
+	return remoteAddr, nil
+}
+
+// InvalidateCache removes a specific endpoint from the address cache
+func (t *HolepunchTester) InvalidateCache(endpoint string) {
+	t.addrCacheMu.Lock()
+	delete(t.addrCache, endpoint)
+	t.addrCacheMu.Unlock()
+}
+
+// ClearCache clears all cached addresses
+func (t *HolepunchTester) ClearCache() {
+	t.addrCacheMu.Lock()
+	t.addrCache = make(map[string]*cachedAddr)
+	t.addrCacheMu.Unlock()
+}
+
 // handleResponse is called by SharedBind when a magic response is received
 func (t *HolepunchTester) handleResponse(addr netip.AddrPort, echoData []byte) {
-	logger.Debug("Received magic response from %s", addr.String())
+	// logger.Debug("Received magic response from %s", addr.String())
 	key := string(echoData)
 
 	value, ok := t.pendingRequests.LoadAndDelete(key)
@@ -152,7 +223,7 @@ func (t *HolepunchTester) handleResponse(addr netip.AddrPort, echoData []byte) {
 
 	req := value.(*pendingRequest)
 	rtt := time.Since(req.sentAt)
-	logger.Debug("Magic response matched pending request for %s (RTT: %v)", req.endpoint, rtt)
+	// logger.Debug("Magic response matched pending request for %s (RTT: %v)", req.endpoint, rtt)
 
 	// Send RTT to the waiting goroutine (non-blocking)
 	select {
@@ -183,20 +254,10 @@ func (t *HolepunchTester) TestEndpoint(endpoint string, timeout time.Duration) T
 		return result
 	}
 
-	// Resolve the endpoint
-	host, err := util.ResolveDomain(endpoint)
+	// Resolve the endpoint (using cache)
+	remoteAddr, err := t.resolveEndpoint(endpoint)
 	if err != nil {
-		host = endpoint
-	}
-
-	_, _, err = net.SplitHostPort(host)
-	if err != nil {
-		host = net.JoinHostPort(host, "21820")
-	}
-
-	remoteAddr, err := net.ResolveUDPAddr("udp", host)
-	if err != nil {
-		result.Error = fmt.Errorf("failed to resolve UDP address %s: %w", host, err)
+		result.Error = err
 		return result
 	}
 

main.go

@@ -1,7 +1,9 @@
 package main
 
 import (
+	"bytes"
 	"context"
+	"crypto/tls"
 	"encoding/json"
 	"errors"
 	"flag"
@@ -11,12 +13,12 @@ import (
 	"net/netip"
 	"os"
 	"os/signal"
-	"path/filepath"
 	"strconv"
 	"strings"
 	"syscall"
 	"time"
 
+	"github.com/fosrl/newt/authdaemon"
 	"github.com/fosrl/newt/docker"
 	"github.com/fosrl/newt/healthcheck"
 	"github.com/fosrl/newt/logger"
@@ -37,6 +39,7 @@ import (
 
 type WgData struct {
 	Endpoint           string               `json:"endpoint"`
+	RelayPort          uint16               `json:"relayPort"`
 	PublicKey          string               `json:"publicKey"`
 	ServerIP           string               `json:"serverIP"`
 	TunnelIP           string               `json:"tunnelIP"`
@@ -57,10 +60,6 @@ type ExitNodeData struct {
 	ExitNodes []ExitNode `json:"exitNodes"`
 }
 
-type SSHPublicKeyData struct {
-	PublicKey string `json:"publicKey"`
-}
-
 // ExitNode represents an exit node with an ID, endpoint, and weight.
 type ExitNode struct {
 	ID                     int     `json:"exitNodeId"`
@@ -116,6 +115,8 @@ var (
 	err                                error
 	logLevel                           string
 	interfaceName                      string
+	port                               uint16
+	portStr                            string
 	disableClients                     bool
 	updownScript                       string
 	dockerSocket                       string
@@ -132,6 +133,11 @@ var (
 	preferEndpoint                     string
 	healthMonitor                      *healthcheck.Monitor
 	enforceHealthcheckCert             bool
+	authDaemonKey                      string
+	authDaemonPrincipalsFile           string
+	authDaemonCACertPath               string
+	authDaemonEnabled                  bool
+	authDaemonGenerateRandomPassword   bool
 	// Build/version (can be overridden via -ldflags "-X main.newtVersion=...")
 	newtVersion = "version_replaceme"
 
@@ -154,10 +160,49 @@ var (
 )
 
 func main() {
+	// Check for subcommands first (only principals exits early)
+	if len(os.Args) > 1 {
+		switch os.Args[1] {
+		case "auth-daemon":
+			// Run principals subcommand only if the next argument is "principals"
+			if len(os.Args) > 2 && os.Args[2] == "principals" {
+				runPrincipalsCmd(os.Args[3:])
+				return
+			}
+
+			// auth-daemon subcommand without "principals" - show help
+			fmt.Println("Error: auth-daemon subcommand requires 'principals' argument")
+			fmt.Println()
+			fmt.Println("Usage:")
+			fmt.Println("  newt auth-daemon principals [options]")
+			fmt.Println()
+
+			// If not "principals", exit the switch to continue with normal execution
+			return
+		}
+	}
+
+	// Check if we're running as a Windows service
+	if isWindowsService() {
+		runService("NewtWireguardService", false, os.Args[1:])
+		return
+	}
+
+	// Handle service management commands on Windows (install, remove, start, stop, etc.)
+	if handleServiceCommand() {
+		return
+	}
+
 	// Prepare context for graceful shutdown and signal handling
 	ctx, stop := signal.NotifyContext(context.Background(), os.Interrupt, syscall.SIGTERM)
 	defer stop()
 
+	// Run the main newt logic
+	runNewtMain(ctx)
+}
+
+// runNewtMain contains the main newt logic, extracted for service support
+func runNewtMain(ctx context.Context) {
 	// if PANGOLIN_ENDPOINT, NEWT_ID, and NEWT_SECRET are set as environment variables, they will be used as default values
 	endpoint = os.Getenv("PANGOLIN_ENDPOINT")
 	id = os.Getenv("NEWT_ID")
@@ -167,6 +212,12 @@ func main() {
 	logLevel = os.Getenv("LOG_LEVEL")
 	updownScript = os.Getenv("UPDOWN_SCRIPT")
 	interfaceName = os.Getenv("INTERFACE")
+	portStr = os.Getenv("PORT")
+	authDaemonKey = os.Getenv("AD_KEY")
+	authDaemonPrincipalsFile = os.Getenv("AD_PRINCIPALS_FILE")
+	authDaemonCACertPath = os.Getenv("AD_CA_CERT_PATH")
+	authDaemonEnabledEnv := os.Getenv("AUTH_DAEMON_ENABLED")
+	authDaemonGenerateRandomPasswordEnv := os.Getenv("AD_GENERATE_RANDOM_PASSWORD")
 
 	// Metrics/observability env mirrors
 	metricsEnabledEnv := os.Getenv("NEWT_METRICS_PROMETHEUS_ENABLED")
@@ -235,6 +286,9 @@ func main() {
 	if interfaceName == "" {
 		flag.StringVar(&interfaceName, "interface", "newt", "Name of the WireGuard interface")
 	}
+	if portStr == "" {
+		flag.StringVar(&portStr, "port", "", "Port for client WireGuard interface")
+	}
 	if useNativeInterfaceEnv == "" {
 		flag.BoolVar(&useNativeInterface, "native", false, "Use native WireGuard interface")
 	}
@@ -256,10 +310,6 @@ func main() {
 	// load the prefer endpoint just as a flag
 	flag.StringVar(&preferEndpoint, "prefer-endpoint", "", "Prefer this endpoint for the connection (if set, will override the endpoint from the server)")
 
-	// if authorizedKeysFile == "" {
-	// 	flag.StringVar(&authorizedKeysFile, "authorized-keys-file", "~/.ssh/authorized_keys", "Path to authorized keys file (if unset, no keys will be authorized)")
-	// }
-
 	// Add new mTLS flags
 	if tlsClientCert == "" {
 		flag.StringVar(&tlsClientCert, "tls-client-cert-file", "", "Path to client certificate file (PEM/DER format)")
@@ -312,7 +362,7 @@ func main() {
 
 	// Metrics/observability flags (mirror ENV if unset)
 	if metricsEnabledEnv == "" {
-		flag.BoolVar(&metricsEnabled, "metrics", true, "Enable Prometheus /metrics exporter")
+		flag.BoolVar(&metricsEnabled, "metrics", false, "Enable Prometheus metrics exporter")
 	} else {
 		if v, err := strconv.ParseBool(metricsEnabledEnv); err == nil {
 			metricsEnabled = v
@@ -347,6 +397,31 @@ func main() {
 		region = regionEnv
 	}
 
+	// Auth daemon flags
+	if authDaemonKey == "" {
+		flag.StringVar(&authDaemonKey, "ad-pre-shared-key", "", "Pre-shared key for auth daemon authentication")
+	}
+	if authDaemonPrincipalsFile == "" {
+		flag.StringVar(&authDaemonPrincipalsFile, "ad-principals-file", "/var/run/auth-daemon/principals", "Path to the principals file for auth daemon")
+	}
+	if authDaemonCACertPath == "" {
+		flag.StringVar(&authDaemonCACertPath, "ad-ca-cert-path", "/etc/ssh/ca.pem", "Path to the CA certificate file for auth daemon")
+	}
+	if authDaemonEnabledEnv == "" {
+		flag.BoolVar(&authDaemonEnabled, "auth-daemon", false, "Enable auth daemon mode (runs alongside normal newt operation)")
+	} else {
+		if v, err := strconv.ParseBool(authDaemonEnabledEnv); err == nil {
+			authDaemonEnabled = v
+		}
+	}
+	if authDaemonGenerateRandomPasswordEnv == "" {
+		flag.BoolVar(&authDaemonGenerateRandomPassword, "ad-generate-random-password", false, "Generate a random password for authenticated users")
+	} else {
+		if v, err := strconv.ParseBool(authDaemonGenerateRandomPasswordEnv); err == nil {
+			authDaemonGenerateRandomPassword = v
+		}
+	}
+
 	// do a --version check
 	version := flag.Bool("version", false, "Print the version")
 
@@ -357,8 +432,31 @@ func main() {
 		tlsClientCAs = append(tlsClientCAs, tlsClientCAsFlag...)
 	}
 
+	if portStr != "" {
+		portInt, err := strconv.Atoi(portStr)
+		if err != nil {
+			logger.Warn("Failed to parse PORT, choosing a random port")
+		} else {
+			port = uint16(portInt)
+		}
+	}
+
+	if *version {
+		fmt.Println("Newt version " + newtVersion)
+		os.Exit(0)
+	} else {
+		logger.Info("Newt version %s", newtVersion)
+	}
+
 	logger.Init(nil)
 	loggerLevel := util.ParseLogLevel(logLevel)
+
+	// Start auth daemon if enabled
+	if authDaemonEnabled {
+		if err := startAuthDaemon(ctx); err != nil {
+			logger.Fatal("Failed to start auth daemon: %v", err)
+		}
+	}
 	logger.GetLogger().SetLevel(loggerLevel)
 
 	// Initialize telemetry after flags are parsed (so flags override env)
@@ -381,7 +479,7 @@ func main() {
 	}
 	if tel != nil {
 		// Admin HTTP server (exposes /metrics when Prometheus exporter is enabled)
-		logger.Info("Starting metrics server on %s", tcfg.AdminAddr)
+		logger.Debug("Starting metrics server on %s", tcfg.AdminAddr)
 		mux := http.NewServeMux()
 		mux.HandleFunc("/healthz", func(w http.ResponseWriter, r *http.Request) { w.WriteHeader(200) })
 		if tel.PrometheusHandler != nil {
@@ -408,13 +506,6 @@ func main() {
 		defer func() { _ = tel.Shutdown(context.Background()) }()
 	}
 
-	if *version {
-		fmt.Println("Newt version " + newtVersion)
-		os.Exit(0)
-	} else {
-		logger.Info("Newt version %s", newtVersion)
-	}
-
 	if err := updates.CheckForUpdate("fosrl", "newt", newtVersion); err != nil {
 		logger.Error("Error checking for updates: %v\n", err)
 	}
@@ -527,6 +618,8 @@ func main() {
 	var connected bool
 	var wgData WgData
 	var dockerEventMonitor *docker.EventMonitor
+	
+	logger.Debug("++++++++++++++++++++++ the port is %d", port)
 
 	if !disableClients {
 		setupClients(client)
@@ -641,7 +734,7 @@ func main() {
 		// Create WireGuard device
 		dev = device.NewDevice(tun, conn.NewDefaultBind(), device.NewLogger(
 			util.MapToWireGuardLogLevel(loggerLevel),
-			"wireguard: ",
+			"gerbil-wireguard: ",
 		))
 
 		host, _, err := net.SplitHostPort(wgData.Endpoint)
@@ -660,7 +753,12 @@ func main() {
 			return
 		}
 
-		clientsHandleNewtConnection(wgData.PublicKey, endpoint)
+		relayPort := wgData.RelayPort
+		if relayPort == 0 {
+			relayPort = 21820
+		}
+
+		clientsHandleNewtConnection(wgData.PublicKey, endpoint, relayPort)
 
 		// Configure WireGuard
 		config := fmt.Sprintf(`private_key=%s
@@ -1131,94 +1229,6 @@ persistent_keepalive_interval=5`, util.FixKey(privateKey.String()), util.FixKey(
 		}
 	})
 
-	// EXPERIMENTAL: WHAT SHOULD WE DO ABOUT SECURITY?
-	client.RegisterHandler("newt/send/ssh/publicKey", func(msg websocket.WSMessage) {
-		logger.Debug("Received SSH public key request")
-
-		var sshPublicKeyData SSHPublicKeyData
-
-		jsonData, err := json.Marshal(msg.Data)
-		if err != nil {
-			logger.Info(fmtErrMarshaling, err)
-			return
-		}
-		if err := json.Unmarshal(jsonData, &sshPublicKeyData); err != nil {
-			logger.Info("Error unmarshaling SSH public key data: %v", err)
-			return
-		}
-
-		sshPublicKey := sshPublicKeyData.PublicKey
-
-		if authorizedKeysFile == "" {
-			logger.Debug("No authorized keys file set, skipping public key response")
-			return
-		}
-
-		// Expand tilde to home directory if present
-		expandedPath := authorizedKeysFile
-		if strings.HasPrefix(authorizedKeysFile, "~/") {
-			homeDir, err := os.UserHomeDir()
-			if err != nil {
-				logger.Error("Failed to get user home directory: %v", err)
-				return
-			}
-			expandedPath = filepath.Join(homeDir, authorizedKeysFile[2:])
-		}
-
-		// if it is set but the file does not exist, create it
-		if _, err := os.Stat(expandedPath); os.IsNotExist(err) {
-			logger.Debug("Authorized keys file does not exist, creating it: %s", expandedPath)
-			if err := os.MkdirAll(filepath.Dir(expandedPath), 0755); err != nil {
-				logger.Error("Failed to create directory for authorized keys file: %v", err)
-				return
-			}
-			if _, err := os.Create(expandedPath); err != nil {
-				logger.Error("Failed to create authorized keys file: %v", err)
-				return
-			}
-		}
-
-		// Check if the public key already exists in the file
-		fileContent, err := os.ReadFile(expandedPath)
-		if err != nil {
-			logger.Error("Failed to read authorized keys file: %v", err)
-			return
-		}
-
-		// Check if the key already exists (trim whitespace for comparison)
-		existingKeys := strings.Split(string(fileContent), "\n")
-		keyAlreadyExists := false
-		trimmedNewKey := strings.TrimSpace(sshPublicKey)
-
-		for _, existingKey := range existingKeys {
-			if strings.TrimSpace(existingKey) == trimmedNewKey && trimmedNewKey != "" {
-				keyAlreadyExists = true
-				break
-			}
-		}
-
-		if keyAlreadyExists {
-			logger.Info("SSH public key already exists in authorized keys file, skipping")
-			return
-		}
-
-		// append the public key to the authorized keys file
-		logger.Debug("Appending public key to authorized keys file: %s", sshPublicKey)
-		file, err := os.OpenFile(expandedPath, os.O_APPEND|os.O_WRONLY, 0644)
-		if err != nil {
-			logger.Error("Failed to open authorized keys file: %v", err)
-			return
-		}
-		defer file.Close()
-
-		if _, err := file.WriteString(sshPublicKey + "\n"); err != nil {
-			logger.Error("Failed to write public key to authorized keys file: %v", err)
-			return
-		}
-
-		logger.Info("SSH public key appended to authorized keys file")
-	})
-
 	// Register handler for adding health check targets
 	client.RegisterHandler("newt/healthcheck/add", func(msg websocket.WSMessage) {
 		logger.Debug("Received health check add request: %+v", msg)
@@ -1374,6 +1384,175 @@ persistent_keepalive_interval=5`, util.FixKey(privateKey.String()), util.FixKey(
 		}
 	})
 
+	// Register handler for SSH certificate issued events
+	client.RegisterHandler("newt/pam/connection", func(msg websocket.WSMessage) {
+		logger.Debug("Received SSH certificate issued message")
+
+		// Define the structure of the incoming message
+		type SSHCertData struct {
+			MessageId          int    `json:"messageId"`
+			AgentPort          int    `json:"agentPort"`
+			AgentHost          string `json:"agentHost"`
+			ExternalAuthDaemon bool   `json:"externalAuthDaemon"`
+			CACert             string `json:"caCert"`
+			Username           string `json:"username"`
+			NiceID             string `json:"niceId"`
+			Metadata           struct {
+				SudoMode     string   `json:"sudoMode"`
+				SudoCommands []string `json:"sudoCommands"`
+				Homedir      bool     `json:"homedir"`
+				Groups       []string `json:"groups"`
+			} `json:"metadata"`
+		}
+
+		var certData SSHCertData
+		jsonData, err := json.Marshal(msg.Data)
+		if err != nil {
+			logger.Error("Error marshaling SSH cert data: %v", err)
+			return
+		}
+
+		// print the received data for debugging
+		logger.Debug("Received SSH cert data: %s", string(jsonData))
+
+		if err := json.Unmarshal(jsonData, &certData); err != nil {
+			logger.Error("Error unmarshaling SSH cert data: %v", err)
+			return
+		}
+
+		// Check if we're running the auth daemon internally
+		if authDaemonServer != nil && !certData.ExternalAuthDaemon { // if the auth daemon is running internally and the external auth daemon is not enabled
+			// Call ProcessConnection directly when running internally
+			logger.Debug("Calling internal auth daemon ProcessConnection for user %s", certData.Username)
+
+			authDaemonServer.ProcessConnection(authdaemon.ConnectionRequest{
+				CaCert:   certData.CACert,
+				NiceId:   certData.NiceID,
+				Username: certData.Username,
+				Metadata: authdaemon.ConnectionMetadata{
+					SudoMode:     certData.Metadata.SudoMode,
+					SudoCommands: certData.Metadata.SudoCommands,
+					Homedir:      certData.Metadata.Homedir,
+					Groups:       certData.Metadata.Groups,
+				},
+			})
+
+			// Send success response back to cloud
+			err = client.SendMessage("ws/round-trip/complete", map[string]interface{}{
+				"messageId": certData.MessageId,
+				"complete":  true,
+			})
+
+			logger.Info("Successfully processed connection via internal auth daemon for user %s", certData.Username)
+		} else {
+			// External auth daemon mode - make HTTP request
+			// Check if auth daemon key is configured
+			if authDaemonKey == "" {
+				logger.Error("Auth daemon key not configured, cannot communicate with daemon")
+				// Send failure response back to cloud
+				err := client.SendMessage("ws/round-trip/complete", map[string]interface{}{
+					"messageId": certData.MessageId,
+					"complete":  true,
+					"error":     "auth daemon key not configured",
+				})
+				if err != nil {
+					logger.Error("Failed to send SSH cert failure response: %v", err)
+				}
+				return
+			}
+
+			// Prepare the request body for the auth daemon
+			requestBody := map[string]interface{}{
+				"caCert":   certData.CACert,
+				"niceId":   certData.NiceID,
+				"username": certData.Username,
+				"metadata": map[string]interface{}{
+					"sudoMode":     certData.Metadata.SudoMode,
+					"sudoCommands": certData.Metadata.SudoCommands,
+					"homedir":      certData.Metadata.Homedir,
+					"groups":       certData.Metadata.Groups,
+				},
+			}
+
+			requestJSON, err := json.Marshal(requestBody)
+			if err != nil {
+				logger.Error("Failed to marshal auth daemon request: %v", err)
+				// Send failure response
+				client.SendMessage("ws/round-trip/complete", map[string]interface{}{
+					"messageId": certData.MessageId,
+					"complete":  true,
+					"error":     fmt.Sprintf("failed to marshal request: %v", err),
+				})
+				return
+			}
+
+			// Create HTTPS client that skips certificate verification
+			// (auth daemon uses self-signed cert)
+			httpClient := &http.Client{
+				Transport: &http.Transport{
+					TLSClientConfig: &tls.Config{
+						InsecureSkipVerify: true,
+					},
+				},
+				Timeout: 10 * time.Second,
+			}
+
+			// Make the request to the auth daemon
+			url := fmt.Sprintf("https://%s:%d/connection", certData.AgentHost, certData.AgentPort)
+			req, err := http.NewRequest("POST", url, bytes.NewBuffer(requestJSON))
+			if err != nil {
+				logger.Error("Failed to create auth daemon request: %v", err)
+				client.SendMessage("ws/round-trip/complete", map[string]interface{}{
+					"messageId": certData.MessageId,
+					"complete":  true,
+					"error":     fmt.Sprintf("failed to create request: %v", err),
+				})
+				return
+			}
+
+			// Set headers
+			req.Header.Set("Content-Type", "application/json")
+			req.Header.Set("Authorization", "Bearer "+authDaemonKey)
+
+			logger.Debug("Sending SSH cert to auth daemon at %s", url)
+
+			// Send the request
+			resp, err := httpClient.Do(req)
+			if err != nil {
+				logger.Error("Failed to connect to auth daemon: %v", err)
+				client.SendMessage("ws/round-trip/complete", map[string]interface{}{
+					"messageId": certData.MessageId,
+					"complete":  true,
+					"error":     fmt.Sprintf("failed to connect to auth daemon: %v", err),
+				})
+				return
+			}
+			defer resp.Body.Close()
+
+			// Check response status
+			if resp.StatusCode != http.StatusOK {
+				logger.Error("Auth daemon returned non-OK status: %d", resp.StatusCode)
+				client.SendMessage("ws/round-trip/complete", map[string]interface{}{
+					"messageId": certData.MessageId,
+					"complete":  true,
+					"error":     fmt.Sprintf("auth daemon returned status %d", resp.StatusCode),
+				})
+				return
+			}
+
+			logger.Info("Successfully registered SSH certificate with external auth daemon for user %s", certData.Username)
+		}
+
+		// Send success response back to cloud
+		err = client.SendMessage("ws/round-trip/complete", map[string]interface{}{
+			"messageId": certData.MessageId,
+			"complete":  true,
+		})
+		if err != nil {
+			logger.Error("Failed to send SSH cert success response: %v", err)
+		}
+	})
+
 	client.OnConnect(func() error {
 		publicKey = privateKey.PublicKey()
 		logger.Debug("Public key: %s", publicKey)
@@ -1389,7 +1568,12 @@ persistent_keepalive_interval=5`, util.FixKey(privateKey.String()), util.FixKey(
 				"noCloud": noCloud,
 			}, 3*time.Second)
 			logger.Debug("Requesting exit nodes from server")
-			clientsOnConnect()
+
+			if client.GetServerVersion() != "" { // to prevent issues with running newt > 1.7 versions with older servers
+				clientsOnConnect()
+			} else {
+				logger.Warn("CLIENTS WILL NOT WORK ON THIS VERSION OF NEWT WITH THIS VERSION OF PANGOLIN, PLEASE UPDATE THE SERVER TO 1.13 OR HIGHER OR DOWNGRADE NEWT")
+			}
 		}
 
 		// Send registration message to the server for backward compatibility
@@ -1443,10 +1627,8 @@ persistent_keepalive_interval=5`, util.FixKey(privateKey.String()), util.FixKey(
 		}
 	}
 
-	// Wait for interrupt signal
-	sigCh := make(chan os.Signal, 1)
-	signal.Notify(sigCh, syscall.SIGINT, syscall.SIGTERM)
-	<-sigCh
+	// Wait for context cancellation (from signal or service stop)
+	<-ctx.Done()
 
 	// Close clients first (including WGTester)
 	closeClients()
@@ -1471,7 +1653,20 @@ persistent_keepalive_interval=5`, util.FixKey(privateKey.String()), util.FixKey(
 		client.Close()
 	}
 	logger.Info("Exiting...")
-	os.Exit(0)
+}
+
+// runNewtMainWithArgs is used by the Windows service to run newt with specific arguments
+// It sets os.Args and then calls runNewtMain
+func runNewtMainWithArgs(ctx context.Context, args []string) {
+	// Set os.Args to include the program name plus the provided args
+	// This allows flag parsing to work correctly
+	os.Args = append([]string{os.Args[0]}, args...)
+
+	// Setup Windows logging if running as a service
+	setupWindowsEventLog()
+
+	// Run the main newt logic
+	runNewtMain(ctx)
 }
 
 // validateTLSConfig validates the TLS configuration

netstack2/handlers.go

@@ -10,12 +10,18 @@ import (
 	"fmt"
 	"io"
 	"net"
+	"net/netip"
+	"os/exec"
 	"sync"
 	"time"
 
 	"github.com/fosrl/newt/logger"
+	"golang.org/x/net/icmp"
+	"golang.org/x/net/ipv4"
 	"gvisor.dev/gvisor/pkg/tcpip"
 	"gvisor.dev/gvisor/pkg/tcpip/adapters/gonet"
+	"gvisor.dev/gvisor/pkg/tcpip/checksum"
+	"gvisor.dev/gvisor/pkg/tcpip/header"
 	"gvisor.dev/gvisor/pkg/tcpip/stack"
 	"gvisor.dev/gvisor/pkg/tcpip/transport/tcp"
 	"gvisor.dev/gvisor/pkg/tcpip/transport/udp"
@@ -58,6 +64,9 @@ const (
 
 	// Buffer size for copying data
 	bufferSize = 32 * 1024
+
+	// icmpTimeout is the default timeout for ICMP ping requests.
+	icmpTimeout = 5 * time.Second
 )
 
 // TCPHandler handles TCP connections from netstack
@@ -72,6 +81,12 @@ type UDPHandler struct {
 	proxyHandler *ProxyHandler
 }
 
+// ICMPHandler handles ICMP packets from netstack
+type ICMPHandler struct {
+	stack        *stack.Stack
+	proxyHandler *ProxyHandler
+}
+
 // NewTCPHandler creates a new TCP handler
 func NewTCPHandler(s *stack.Stack, ph *ProxyHandler) *TCPHandler {
 	return &TCPHandler{stack: s, proxyHandler: ph}
@@ -82,6 +97,11 @@ func NewUDPHandler(s *stack.Stack, ph *ProxyHandler) *UDPHandler {
 	return &UDPHandler{stack: s, proxyHandler: ph}
 }
 
+// NewICMPHandler creates a new ICMP handler
+func NewICMPHandler(s *stack.Stack, ph *ProxyHandler) *ICMPHandler {
+	return &ICMPHandler{stack: s, proxyHandler: ph}
+}
+
 // InstallTCPHandler installs the TCP forwarder on the stack
 func (h *TCPHandler) InstallTCPHandler() error {
 	tcpForwarder := tcp.NewForwarder(h.stack, defaultWndSize, maxConnAttempts, func(r *tcp.ForwarderRequest) {
@@ -348,3 +368,334 @@ func copyPacketData(dst, src net.PacketConn, to net.Addr, timeout time.Duration)
 		dst.SetReadDeadline(time.Now().Add(timeout))
 	}
 }
+
+// InstallICMPHandler installs the ICMP handler on the stack
+func (h *ICMPHandler) InstallICMPHandler() error {
+	h.stack.SetTransportProtocolHandler(header.ICMPv4ProtocolNumber, h.handleICMPPacket)
+	logger.Debug("ICMP Handler: Installed ICMP protocol handler")
+	return nil
+}
+
+// handleICMPPacket handles incoming ICMP packets
+func (h *ICMPHandler) handleICMPPacket(id stack.TransportEndpointID, pkt *stack.PacketBuffer) bool {
+	logger.Debug("ICMP Handler: Received ICMP packet from %s to %s", id.RemoteAddress, id.LocalAddress)
+
+	// Get the ICMP header from the packet
+	icmpData := pkt.TransportHeader().Slice()
+	if len(icmpData) < header.ICMPv4MinimumSize {
+		logger.Debug("ICMP Handler: Packet too small for ICMP header: %d bytes", len(icmpData))
+		return false
+	}
+
+	icmpHdr := header.ICMPv4(icmpData)
+	icmpType := icmpHdr.Type()
+	icmpCode := icmpHdr.Code()
+
+	logger.Debug("ICMP Handler: Type=%d, Code=%d, Ident=%d, Seq=%d",
+		icmpType, icmpCode, icmpHdr.Ident(), icmpHdr.Sequence())
+
+	// Only handle Echo Request (ping)
+	if icmpType != header.ICMPv4Echo {
+		logger.Debug("ICMP Handler: Ignoring non-echo ICMP type: %d", icmpType)
+		return false
+	}
+
+	// Extract source and destination addresses
+	srcIP := id.RemoteAddress.String()
+	dstIP := id.LocalAddress.String()
+
+	logger.Info("ICMP Handler: Echo Request from %s to %s (ident=%d, seq=%d)",
+		srcIP, dstIP, icmpHdr.Ident(), icmpHdr.Sequence())
+
+	// Convert to netip.Addr for subnet matching
+	srcAddr, err := netip.ParseAddr(srcIP)
+	if err != nil {
+		logger.Debug("ICMP Handler: Failed to parse source IP %s: %v", srcIP, err)
+		return false
+	}
+	dstAddr, err := netip.ParseAddr(dstIP)
+	if err != nil {
+		logger.Debug("ICMP Handler: Failed to parse dest IP %s: %v", dstIP, err)
+		return false
+	}
+
+	// Check subnet rules (use port 0 for ICMP since it doesn't have ports)
+	if h.proxyHandler == nil {
+		logger.Debug("ICMP Handler: No proxy handler configured")
+		return false
+	}
+
+	matchedRule := h.proxyHandler.subnetLookup.Match(srcAddr, dstAddr, 0, header.ICMPv4ProtocolNumber)
+	if matchedRule == nil {
+		logger.Debug("ICMP Handler: No matching subnet rule for %s -> %s", srcIP, dstIP)
+		return false
+	}
+
+	logger.Info("ICMP Handler: Matched subnet rule for %s -> %s", srcIP, dstIP)
+
+	// Determine actual destination (with possible rewrite)
+	actualDstIP := dstIP
+	if matchedRule.RewriteTo != "" {
+		resolvedAddr, err := h.proxyHandler.resolveRewriteAddress(matchedRule.RewriteTo)
+		if err != nil {
+			logger.Info("ICMP Handler: Failed to resolve rewrite address %s: %v", matchedRule.RewriteTo, err)
+		} else {
+			actualDstIP = resolvedAddr.String()
+			logger.Info("ICMP Handler: Using rewritten destination %s (original: %s)", actualDstIP, dstIP)
+		}
+	}
+
+	// Get the full ICMP payload (including the data after the header)
+	icmpPayload := pkt.Data().AsRange().ToSlice()
+
+	// Handle the ping in a goroutine to avoid blocking
+	go h.proxyPing(srcIP, dstIP, actualDstIP, icmpHdr.Ident(), icmpHdr.Sequence(), icmpPayload)
+
+	return true
+}
+
+// proxyPing sends a ping to the actual destination and injects the reply back
+func (h *ICMPHandler) proxyPing(srcIP, originalDstIP, actualDstIP string, ident, seq uint16, payload []byte) {
+	logger.Debug("ICMP Handler: Proxying ping from %s to %s (actual: %s), ident=%d, seq=%d",
+		srcIP, originalDstIP, actualDstIP, ident, seq)
+
+	// Try three methods in order: ip4:icmp -> udp4 -> ping command
+	// Track which method succeeded so we can handle identifier matching correctly
+	method, success := h.tryICMPMethods(actualDstIP, ident, seq, payload)
+
+	if !success {
+		logger.Info("ICMP Handler: All ping methods failed for %s", actualDstIP)
+		return
+	}
+
+	logger.Info("ICMP Handler: Ping successful to %s using %s, injecting reply (ident=%d, seq=%d)",
+		actualDstIP, method, ident, seq)
+
+	// Build the reply packet to inject back into the netstack
+	// The reply should appear to come from the original destination (before rewrite)
+	h.injectICMPReply(srcIP, originalDstIP, ident, seq, payload)
+}
+
+// tryICMPMethods tries all available ICMP methods in order
+func (h *ICMPHandler) tryICMPMethods(actualDstIP string, ident, seq uint16, payload []byte) (string, bool) {
+	if h.tryRawICMP(actualDstIP, ident, seq, payload, false) {
+		return "raw ICMP", true
+	}
+	if h.tryUnprivilegedICMP(actualDstIP, ident, seq, payload) {
+		return "unprivileged ICMP", true
+	}
+	if h.tryPingCommand(actualDstIP, ident, seq, payload) {
+		return "ping command", true
+	}
+	return "", false
+}
+
+// tryRawICMP attempts to ping using raw ICMP sockets (requires CAP_NET_RAW or root)
+func (h *ICMPHandler) tryRawICMP(actualDstIP string, ident, seq uint16, payload []byte, ignoreIdent bool) bool {
+	conn, err := icmp.ListenPacket("ip4:icmp", "0.0.0.0")
+	if err != nil {
+		logger.Debug("ICMP Handler: Raw ICMP socket not available: %v", err)
+		return false
+	}
+	defer conn.Close()
+
+	logger.Debug("ICMP Handler: Using raw ICMP socket")
+	return h.sendAndReceiveICMP(conn, actualDstIP, ident, seq, payload, false, ignoreIdent)
+}
+
+// tryUnprivilegedICMP attempts to ping using unprivileged ICMP (requires ping_group_range configured)
+func (h *ICMPHandler) tryUnprivilegedICMP(actualDstIP string, ident, seq uint16, payload []byte) bool {
+	conn, err := icmp.ListenPacket("udp4", "0.0.0.0")
+	if err != nil {
+		logger.Debug("ICMP Handler: Unprivileged ICMP socket not available: %v", err)
+		return false
+	}
+	defer conn.Close()
+
+	logger.Debug("ICMP Handler: Using unprivileged ICMP socket")
+	// Unprivileged ICMP doesn't let us control the identifier, so we ignore it in matching
+	return h.sendAndReceiveICMP(conn, actualDstIP, ident, seq, payload, true, true)
+}
+
+// sendAndReceiveICMP sends an ICMP echo request and waits for the reply
+func (h *ICMPHandler) sendAndReceiveICMP(conn *icmp.PacketConn, actualDstIP string, ident, seq uint16, payload []byte, isUnprivileged bool, ignoreIdent bool) bool {
+	// Build the ICMP echo request message
+	echoMsg := &icmp.Message{
+		Type: ipv4.ICMPTypeEcho,
+		Code: 0,
+		Body: &icmp.Echo{
+			ID:   int(ident),
+			Seq:  int(seq),
+			Data: payload,
+		},
+	}
+
+	msgBytes, err := echoMsg.Marshal(nil)
+	if err != nil {
+		logger.Debug("ICMP Handler: Failed to marshal ICMP message: %v", err)
+		return false
+	}
+
+	// Resolve destination address based on socket type
+	var writeErr error
+	if isUnprivileged {
+		// For unprivileged ICMP, use UDP-style addressing
+		udpAddr := &net.UDPAddr{IP: net.ParseIP(actualDstIP)}
+		logger.Debug("ICMP Handler: Sending ping to %s (unprivileged)", udpAddr.String())
+		conn.SetDeadline(time.Now().Add(icmpTimeout))
+		_, writeErr = conn.WriteTo(msgBytes, udpAddr)
+	} else {
+		// For raw ICMP, use IP addressing
+		dst, err := net.ResolveIPAddr("ip4", actualDstIP)
+		if err != nil {
+			logger.Debug("ICMP Handler: Failed to resolve destination %s: %v", actualDstIP, err)
+			return false
+		}
+		logger.Debug("ICMP Handler: Sending ping to %s (raw)", dst.String())
+		conn.SetDeadline(time.Now().Add(icmpTimeout))
+		_, writeErr = conn.WriteTo(msgBytes, dst)
+	}
+
+	if writeErr != nil {
+		logger.Debug("ICMP Handler: Failed to send ping to %s: %v", actualDstIP, writeErr)
+		return false
+	}
+
+	logger.Debug("ICMP Handler: Ping sent to %s, waiting for reply (ident=%d, seq=%d)", actualDstIP, ident, seq)
+
+	// Wait for reply - loop to filter out non-matching packets
+	replyBuf := make([]byte, 1500)
+
+	for {
+		n, peer, err := conn.ReadFrom(replyBuf)
+		if err != nil {
+			logger.Debug("ICMP Handler: Failed to receive ping reply from %s: %v", actualDstIP, err)
+			return false
+		}
+
+		logger.Debug("ICMP Handler: Received %d bytes from %s", n, peer.String())
+
+		// Parse the reply
+		replyMsg, err := icmp.ParseMessage(1, replyBuf[:n])
+		if err != nil {
+			logger.Debug("ICMP Handler: Failed to parse ICMP message: %v", err)
+			continue
+		}
+
+		// Check if it's an echo reply (type 0), not an echo request (type 8)
+		if replyMsg.Type != ipv4.ICMPTypeEchoReply {
+			logger.Debug("ICMP Handler: Received non-echo-reply type: %v, continuing to wait", replyMsg.Type)
+			continue
+		}
+
+		reply, ok := replyMsg.Body.(*icmp.Echo)
+		if !ok {
+			logger.Debug("ICMP Handler: Invalid echo reply body type, continuing to wait")
+			continue
+		}
+
+		// Verify the sequence matches what we sent
+		// For unprivileged ICMP, the kernel controls the identifier, so we only check sequence
+		if reply.Seq != int(seq) {
+			logger.Debug("ICMP Handler: Reply seq mismatch: got seq=%d, want seq=%d", reply.Seq, seq)
+			continue
+		}
+
+		if !ignoreIdent && reply.ID != int(ident) {
+			logger.Debug("ICMP Handler: Reply ident mismatch: got ident=%d, want ident=%d", reply.ID, ident)
+			continue
+		}
+
+		// Found matching reply
+		logger.Debug("ICMP Handler: Received valid echo reply")
+		return true
+	}
+}
+
+// tryPingCommand attempts to ping using the system ping command (always works, but less control)
+func (h *ICMPHandler) tryPingCommand(actualDstIP string, ident, seq uint16, payload []byte) bool {
+	logger.Debug("ICMP Handler: Attempting to use system ping command")
+
+	ctx, cancel := context.WithTimeout(context.Background(), icmpTimeout)
+	defer cancel()
+
+	// Send one ping with timeout
+	// -c 1: count = 1 packet
+	// -W 5: timeout = 5 seconds
+	// -q: quiet output (just summary)
+	cmd := exec.CommandContext(ctx, "ping", "-c", "1", "-W", "5", "-q", actualDstIP)
+	output, err := cmd.CombinedOutput()
+
+	if err != nil {
+		logger.Debug("ICMP Handler: System ping command failed: %v, output: %s", err, string(output))
+		return false
+	}
+
+	logger.Debug("ICMP Handler: System ping command succeeded")
+	return true
+}
+
+// injectICMPReply creates an ICMP echo reply packet and queues it to be sent back through the tunnel
+func (h *ICMPHandler) injectICMPReply(dstIP, srcIP string, ident, seq uint16, payload []byte) {
+	logger.Debug("ICMP Handler: Creating reply from %s to %s (ident=%d, seq=%d)",
+		srcIP, dstIP, ident, seq)
+
+	// Parse addresses
+	srcAddr, err := netip.ParseAddr(srcIP)
+	if err != nil {
+		logger.Info("ICMP Handler: Failed to parse source IP for reply: %v", err)
+		return
+	}
+	dstAddr, err := netip.ParseAddr(dstIP)
+	if err != nil {
+		logger.Info("ICMP Handler: Failed to parse dest IP for reply: %v", err)
+		return
+	}
+
+	// Calculate total packet size
+	ipHeaderLen := header.IPv4MinimumSize
+	icmpHeaderLen := header.ICMPv4MinimumSize
+	totalLen := ipHeaderLen + icmpHeaderLen + len(payload)
+
+	// Create the packet buffer
+	pkt := make([]byte, totalLen)
+
+	// Build IPv4 header
+	ipHdr := header.IPv4(pkt[:ipHeaderLen])
+	ipHdr.Encode(&header.IPv4Fields{
+		TotalLength: uint16(totalLen),
+		TTL:         64,
+		Protocol:    uint8(header.ICMPv4ProtocolNumber),
+		SrcAddr:     tcpip.AddrFrom4(srcAddr.As4()),
+		DstAddr:     tcpip.AddrFrom4(dstAddr.As4()),
+	})
+	ipHdr.SetChecksum(^ipHdr.CalculateChecksum())
+
+	// Build ICMP header
+	icmpHdr := header.ICMPv4(pkt[ipHeaderLen : ipHeaderLen+icmpHeaderLen])
+	icmpHdr.SetType(header.ICMPv4EchoReply)
+	icmpHdr.SetCode(0)
+	icmpHdr.SetIdent(ident)
+	icmpHdr.SetSequence(seq)
+
+	// Copy payload
+	copy(pkt[ipHeaderLen+icmpHeaderLen:], payload)
+
+	// Calculate ICMP checksum (covers ICMP header + payload)
+	icmpHdr.SetChecksum(0)
+	icmpData := pkt[ipHeaderLen:]
+	icmpHdr.SetChecksum(^checksum.Checksum(icmpData, 0))
+
+	logger.Debug("ICMP Handler: Built reply packet, total length=%d", totalLen)
+
+	// Queue the packet to be sent back through the tunnel
+	if h.proxyHandler != nil {
+		if h.proxyHandler.QueueICMPReply(pkt) {
+			logger.Info("ICMP Handler: Queued echo reply packet for transmission")
+		} else {
+			logger.Info("ICMP Handler: Failed to queue echo reply packet")
+		}
+	} else {
+		logger.Info("ICMP Handler: Cannot queue reply - proxy handler not available")
+	}
+}

netstack2/proxy.go

@@ -22,10 +22,12 @@ import (
 	"gvisor.dev/gvisor/pkg/tcpip/transport/udp"
 )
 
-// PortRange represents an allowed range of ports (inclusive)
+// PortRange represents an allowed range of ports (inclusive) with optional protocol filtering
+// Protocol can be "tcp", "udp", or "" (empty string means both protocols)
 type PortRange struct {
-	Min uint16
-	Max uint16
+	Min      uint16
+	Max      uint16
+	Protocol string // "tcp", "udp", or "" for both
 }
 
 // SubnetRule represents a subnet with optional port restrictions and source address
@@ -41,101 +43,11 @@ type PortRange struct {
 type SubnetRule struct {
 	SourcePrefix netip.Prefix // Source IP prefix (who is sending)
 	DestPrefix   netip.Prefix // Destination IP prefix (where it's going)
+	DisableIcmp  bool         // If true, ICMP traffic is blocked for this subnet
 	RewriteTo    string       // Optional rewrite address for DNAT - can be IP/CIDR or domain name
 	PortRanges   []PortRange  // empty slice means all ports allowed
 }
 
-// ruleKey is used as a map key for fast O(1) lookups
-type ruleKey struct {
-	sourcePrefix string
-	destPrefix   string
-}
-
-// SubnetLookup provides fast IP subnet and port matching with O(1) lookup performance
-type SubnetLookup struct {
-	mu    sync.RWMutex
-	rules map[ruleKey]*SubnetRule // Map for O(1) lookups by prefix combination
-}
-
-// NewSubnetLookup creates a new subnet lookup table
-func NewSubnetLookup() *SubnetLookup {
-	return &SubnetLookup{
-		rules: make(map[ruleKey]*SubnetRule),
-	}
-}
-
-// AddSubnet adds a subnet rule with source and destination prefixes and optional port restrictions
-// If portRanges is nil or empty, all ports are allowed for this subnet
-// rewriteTo can be either an IP/CIDR (e.g., "192.168.1.1/32") or a domain name (e.g., "example.com")
-func (sl *SubnetLookup) AddSubnet(sourcePrefix, destPrefix netip.Prefix, rewriteTo string, portRanges []PortRange) {
-	sl.mu.Lock()
-	defer sl.mu.Unlock()
-
-	key := ruleKey{
-		sourcePrefix: sourcePrefix.String(),
-		destPrefix:   destPrefix.String(),
-	}
-
-	sl.rules[key] = &SubnetRule{
-		SourcePrefix: sourcePrefix,
-		DestPrefix:   destPrefix,
-		RewriteTo:    rewriteTo,
-		PortRanges:   portRanges,
-	}
-}
-
-// RemoveSubnet removes a subnet rule from the lookup table
-func (sl *SubnetLookup) RemoveSubnet(sourcePrefix, destPrefix netip.Prefix) {
-	sl.mu.Lock()
-	defer sl.mu.Unlock()
-
-	key := ruleKey{
-		sourcePrefix: sourcePrefix.String(),
-		destPrefix:   destPrefix.String(),
-	}
-
-	delete(sl.rules, key)
-}
-
-// Match checks if a source IP, destination IP, and port match any subnet rule
-// Returns the matched rule if BOTH:
-//   - The source IP is in the rule's source prefix
-//   - The destination IP is in the rule's destination prefix
-//   - The port is in an allowed range (or no port restrictions exist)
-//
-// Returns nil if no rule matches
-func (sl *SubnetLookup) Match(srcIP, dstIP netip.Addr, port uint16) *SubnetRule {
-	sl.mu.RLock()
-	defer sl.mu.RUnlock()
-
-	// Iterate through all rules to find matching source and destination prefixes
-	// This is O(n) but necessary since we need to check prefix containment, not exact match
-	for _, rule := range sl.rules {
-		// Check if source and destination IPs match their respective prefixes
-		if !rule.SourcePrefix.Contains(srcIP) {
-			continue
-		}
-		if !rule.DestPrefix.Contains(dstIP) {
-			continue
-		}
-
-		// Both IPs match - now check port restrictions
-		// If no port ranges specified, all ports are allowed
-		if len(rule.PortRanges) == 0 {
-			return rule
-		}
-
-		// Check if port is in any of the allowed ranges
-		for _, pr := range rule.PortRanges {
-			if port >= pr.Min && port <= pr.Max {
-				return rule
-			}
-		}
-	}
-
-	return nil
-}
-
 // connKey uniquely identifies a connection for NAT tracking
 type connKey struct {
 	srcIP   string
@@ -145,6 +57,17 @@ type connKey struct {
 	proto   uint8
 }
 
+// reverseConnKey uniquely identifies a connection for reverse NAT lookup (reply direction)
+// Key structure: (rewrittenTo, originalSrcIP, originalSrcPort, originalDstPort, proto)
+// This allows O(1) lookup of NAT entries for reply packets
+type reverseConnKey struct {
+	rewrittenTo     string // The address we rewrote to (becomes src in replies)
+	originalSrcIP   string // Original source IP (becomes dst in replies)
+	originalSrcPort uint16 // Original source port (becomes dst port in replies)
+	originalDstPort uint16 // Original destination port (becomes src port in replies)
+	proto           uint8
+}
+
 // destKey identifies a destination for handler lookups (without source port since it may change)
 type destKey struct {
 	srcIP   string
@@ -166,23 +89,28 @@ type ProxyHandler struct {
 	proxyNotifyHandle *channel.NotificationHandle
 	tcpHandler        *TCPHandler
 	udpHandler        *UDPHandler
+	icmpHandler       *ICMPHandler
 	subnetLookup      *SubnetLookup
 	natTable          map[connKey]*natState
-	destRewriteTable  map[destKey]netip.Addr // Maps original dest to rewritten dest for handler lookups
+	reverseNatTable   map[reverseConnKey]*natState // Reverse lookup map for O(1) reply packet NAT
+	destRewriteTable  map[destKey]netip.Addr       // Maps original dest to rewritten dest for handler lookups
 	natMu             sync.RWMutex
 	enabled           bool
+	icmpReplies       chan []byte          // Channel for ICMP reply packets to be sent back through the tunnel
+	notifiable        channel.Notification // Notification handler for triggering reads
 }
 
 // ProxyHandlerOptions configures the proxy handler
 type ProxyHandlerOptions struct {
-	EnableTCP bool
-	EnableUDP bool
-	MTU       int
+	EnableTCP  bool
+	EnableUDP  bool
+	EnableICMP bool
+	MTU        int
 }
 
 // NewProxyHandler creates a new proxy handler for promiscuous mode
 func NewProxyHandler(options ProxyHandlerOptions) (*ProxyHandler, error) {
-	if !options.EnableTCP && !options.EnableUDP {
+	if !options.EnableTCP && !options.EnableUDP && !options.EnableICMP {
 		return nil, nil // No proxy needed
 	}
 
@@ -190,7 +118,9 @@ func NewProxyHandler(options ProxyHandlerOptions) (*ProxyHandler, error) {
 		enabled:          true,
 		subnetLookup:     NewSubnetLookup(),
 		natTable:         make(map[connKey]*natState),
+		reverseNatTable:  make(map[reverseConnKey]*natState),
 		destRewriteTable: make(map[destKey]netip.Addr),
+		icmpReplies:      make(chan []byte, 256), // Buffer for ICMP reply packets
 		proxyEp:          channel.New(1024, uint32(options.MTU), ""),
 		proxyStack: stack.New(stack.Options{
 			NetworkProtocols: []stack.NetworkProtocolFactory{
@@ -222,6 +152,15 @@ func NewProxyHandler(options ProxyHandlerOptions) (*ProxyHandler, error) {
 		}
 	}
 
+	// Initialize ICMP handler if enabled
+	if options.EnableICMP {
+		handler.icmpHandler = NewICMPHandler(handler.proxyStack, handler)
+		if err := handler.icmpHandler.InstallICMPHandler(); err != nil {
+			return nil, fmt.Errorf("failed to install ICMP handler: %v", err)
+		}
+		logger.Debug("ProxyHandler: ICMP handler enabled")
+	}
+
 	// // Example 1: Add a rule with no port restrictions (all ports allowed)
 	// // This accepts all traffic FROM 10.0.0.0/24 TO 10.20.20.0/24
 	// sourceSubnet := netip.MustParsePrefix("10.0.0.0/24")
@@ -246,11 +185,11 @@ func NewProxyHandler(options ProxyHandlerOptions) (*ProxyHandler, error) {
 // destPrefix: The IP prefix of the destination
 // rewriteTo: Optional address to rewrite destination to - can be IP/CIDR or domain name
 // If portRanges is nil or empty, all ports are allowed for this subnet
-func (p *ProxyHandler) AddSubnetRule(sourcePrefix, destPrefix netip.Prefix, rewriteTo string, portRanges []PortRange) {
+func (p *ProxyHandler) AddSubnetRule(sourcePrefix, destPrefix netip.Prefix, rewriteTo string, portRanges []PortRange, disableIcmp bool) {
 	if p == nil || !p.enabled {
 		return
 	}
-	p.subnetLookup.AddSubnet(sourcePrefix, destPrefix, rewriteTo, portRanges)
+	p.subnetLookup.AddSubnet(sourcePrefix, destPrefix, rewriteTo, portRanges, disableIcmp)
 }
 
 // RemoveSubnetRule removes a subnet from the proxy handler
@@ -329,6 +268,9 @@ func (p *ProxyHandler) Initialize(notifiable channel.Notification) error {
 		return nil
 	}
 
+	// Store notifiable for triggering notifications on ICMP replies
+	p.notifiable = notifiable
+
 	// Add notification handler
 	p.proxyNotifyHandle = p.proxyEp.AddNotify(notifiable)
 
@@ -407,14 +349,21 @@ func (p *ProxyHandler) HandleIncomingPacket(packet []byte) bool {
 		}
 		udpHeader := header.UDP(packet[headerLen:])
 		dstPort = udpHeader.DestinationPort()
-	default:
-		// For other protocols (ICMP, etc.), use port 0 (must match rules with no port restrictions)
+	case header.ICMPv4ProtocolNumber:
+		// ICMP doesn't have ports, use port 0 (must match rules with no port restrictions)
 		dstPort = 0
+		logger.Debug("HandleIncomingPacket: ICMP packet from %s to %s", srcAddr, dstAddr)
+	default:
+		// For other protocols, use port 0 (must match rules with no port restrictions)
+		dstPort = 0
+		logger.Debug("HandleIncomingPacket: Unknown protocol %d from %s to %s", protocol, srcAddr, dstAddr)
 	}
 
-	// Check if the source IP, destination IP, and port match any subnet rule
-	matchedRule := p.subnetLookup.Match(srcAddr, dstAddr, dstPort)
+	// Check if the source IP, destination IP, port, and protocol match any subnet rule
+	matchedRule := p.subnetLookup.Match(srcAddr, dstAddr, dstPort, protocol)
 	if matchedRule != nil {
+		logger.Debug("HandleIncomingPacket: Matched rule for %s -> %s (proto=%d, port=%d)",
+			srcAddr, dstAddr, protocol, dstPort)
 		// Check if we need to perform DNAT
 		if matchedRule.RewriteTo != "" {
 			// Create connection tracking key using original destination
@@ -472,10 +421,23 @@ func (p *ProxyHandler) HandleIncomingPacket(packet []byte) bool {
 
 				// Store NAT state for this connection
 				p.natMu.Lock()
-				p.natTable[key] = &natState{
+				natEntry := &natState{
 					originalDst: dstAddr,
 					rewrittenTo: newDst,
 				}
+				p.natTable[key] = natEntry
+
+				// Create reverse lookup key for O(1) reply packet lookups
+				// Key: (rewrittenTo, originalSrcIP, originalSrcPort, originalDstPort, proto)
+				reverseKey := reverseConnKey{
+					rewrittenTo:     newDst.String(),
+					originalSrcIP:   srcAddr.String(),
+					originalSrcPort: srcPort,
+					originalDstPort: dstPort,
+					proto:           uint8(protocol),
+				}
+				p.reverseNatTable[reverseKey] = natEntry
+
 				// Store destination rewrite for handler lookups
 				p.destRewriteTable[dKey] = newDst
 				p.natMu.Unlock()
@@ -501,9 +463,12 @@ func (p *ProxyHandler) HandleIncomingPacket(packet []byte) bool {
 			Payload: buffer.MakeWithData(packet),
 		})
 		p.proxyEp.InjectInbound(header.IPv4ProtocolNumber, pkb)
+		logger.Debug("HandleIncomingPacket: Injected packet into proxy stack (proto=%d)", protocol)
 		return true
 	}
 
+	// logger.Debug("HandleIncomingPacket: No matching rule for %s -> %s (proto=%d, port=%d)",
+		// srcAddr, dstAddr, protocol, dstPort)
 	return false
 }
 
@@ -626,6 +591,15 @@ func (p *ProxyHandler) ReadOutgoingPacket() *buffer.View {
 		return nil
 	}
 
+	// First check for ICMP reply packets (non-blocking)
+	select {
+	case icmpReply := <-p.icmpReplies:
+		logger.Debug("ReadOutgoingPacket: Returning ICMP reply packet (%d bytes)", len(icmpReply))
+		return buffer.NewViewWithData(icmpReply)
+	default:
+		// No ICMP reply available, continue to check proxy endpoint
+	}
+
 	pkt := p.proxyEp.Read()
 	if pkt != nil {
 		view := pkt.ToView()
@@ -655,22 +629,29 @@ func (p *ProxyHandler) ReadOutgoingPacket() *buffer.View {
 					srcPort = udpHeader.SourcePort()
 					dstPort = udpHeader.DestinationPort()
 				}
+			case header.ICMPv4ProtocolNumber:
+				// ICMP packets don't need NAT translation in our implementation
+				// since we construct reply packets with the correct addresses
+				logger.Debug("ReadOutgoingPacket: ICMP packet from %s to %s", srcIP, dstIP)
+				return view
 			}
 
-			// Look up NAT state for reverse translation
-			// The key uses the original dst (before rewrite), so for replies we need to
-			// find the entry where the rewritten address matches the current source
+			// Look up NAT state for reverse translation using O(1) reverse lookup map
+			// Key: (rewrittenTo, originalSrcIP, originalSrcPort, originalDstPort, proto)
+			// For reply packets:
+			// - reply's srcIP = rewrittenTo (the address we rewrote to)
+			// - reply's dstIP = originalSrcIP (original source IP)
+			// - reply's srcPort = originalDstPort (original destination port)
+			// - reply's dstPort = originalSrcPort (original source port)
 			p.natMu.RLock()
-			var natEntry *natState
-			for k, entry := range p.natTable {
-				// Match: reply's dst should be original src, reply's src should be rewritten dst
-				if k.srcIP == dstIP.String() && k.srcPort == dstPort &&
-					entry.rewrittenTo.String() == srcIP.String() && k.dstPort == srcPort &&
-					k.proto == uint8(protocol) {
-					natEntry = entry
-					break
-				}
+			reverseKey := reverseConnKey{
+				rewrittenTo:     srcIP.String(), // Reply's source is the rewritten address
+				originalSrcIP:   dstIP.String(), // Reply's destination is the original source
+				originalSrcPort: dstPort,        // Reply's destination port is the original source port
+				originalDstPort: srcPort,        // Reply's source port is the original destination port
+				proto:           uint8(protocol),
 			}
+			natEntry := p.reverseNatTable[reverseKey]
 			p.natMu.RUnlock()
 
 			if natEntry != nil {
@@ -688,12 +669,37 @@ func (p *ProxyHandler) ReadOutgoingPacket() *buffer.View {
 	return nil
 }
 
+// QueueICMPReply queues an ICMP reply packet to be sent back through the tunnel
+func (p *ProxyHandler) QueueICMPReply(packet []byte) bool {
+	if p == nil || !p.enabled {
+		return false
+	}
+
+	select {
+	case p.icmpReplies <- packet:
+		logger.Debug("QueueICMPReply: Queued ICMP reply packet (%d bytes)", len(packet))
+		// Trigger notification so WriteNotify picks up the packet
+		if p.notifiable != nil {
+			p.notifiable.WriteNotify()
+		}
+		return true
+	default:
+		logger.Info("QueueICMPReply: ICMP reply channel full, dropping packet")
+		return false
+	}
+}
+
 // Close cleans up the proxy handler resources
 func (p *ProxyHandler) Close() error {
 	if p == nil || !p.enabled {
 		return nil
 	}
 
+	// Close ICMP replies channel
+	if p.icmpReplies != nil {
+		close(p.icmpReplies)
+	}
+
 	if p.proxyStack != nil {
 		p.proxyStack.RemoveNIC(1)
 		p.proxyStack.Close()

netstack2/subnet_lookup.go

@@ -0,0 +1,206 @@
+package netstack2
+
+import (
+	"net/netip"
+	"sync"
+
+	"github.com/gaissmai/bart"
+	"gvisor.dev/gvisor/pkg/tcpip"
+	"gvisor.dev/gvisor/pkg/tcpip/header"
+)
+
+// SubnetLookup provides fast IP subnet and port matching using BART (Binary Aggregated Range Tree)
+// This uses BART Table for O(log n) prefix matching with Supernets() for efficient lookups
+//
+// Architecture:
+// - Two-level BART structure for matching both source AND destination prefixes
+// - Level 1: Source prefix -> Level 2 (destination prefix -> rules)
+// - This reduces search space: only check destination prefixes for matching source prefixes
+type SubnetLookup struct {
+	mu sync.RWMutex
+	// Two-level BART structure:
+	// Level 1: Source prefix -> Level 2 (destination prefix -> rules)
+	// This allows us to first match source prefix, then only check destination prefixes
+	// for matching source prefixes, reducing the search space significantly
+	sourceTrie *bart.Table[*destTrie]
+}
+
+// destTrie is a BART for destination prefixes, containing the actual rules
+type destTrie struct {
+	trie  *bart.Table[[]*SubnetRule]
+	rules []*SubnetRule // All rules for this source prefix (for iteration if needed)
+}
+
+// NewSubnetLookup creates a new subnet lookup table using BART
+func NewSubnetLookup() *SubnetLookup {
+	return &SubnetLookup{
+		sourceTrie: &bart.Table[*destTrie]{},
+	}
+}
+
+// prefixEqual compares two prefixes after masking to handle host bits correctly.
+// For example, 10.0.0.5/24 and 10.0.0.0/24 are treated as equal.
+func prefixEqual(a, b netip.Prefix) bool {
+	return a.Masked() == b.Masked()
+}
+
+// AddSubnet adds a subnet rule with source and destination prefixes and optional port restrictions
+// If portRanges is nil or empty, all ports are allowed for this subnet
+// rewriteTo can be either an IP/CIDR (e.g., "192.168.1.1/32") or a domain name (e.g., "example.com")
+func (sl *SubnetLookup) AddSubnet(sourcePrefix, destPrefix netip.Prefix, rewriteTo string, portRanges []PortRange, disableIcmp bool) {
+	sl.mu.Lock()
+	defer sl.mu.Unlock()
+
+	rule := &SubnetRule{
+		SourcePrefix: sourcePrefix,
+		DestPrefix:   destPrefix,
+		DisableIcmp:  disableIcmp,
+		RewriteTo:    rewriteTo,
+		PortRanges:   portRanges,
+	}
+
+	// Canonicalize source prefix to handle host bits correctly
+	canonicalSourcePrefix := sourcePrefix.Masked()
+
+	// Get or create destination trie for this source prefix
+	destTriePtr, exists := sl.sourceTrie.Get(canonicalSourcePrefix)
+	if !exists {
+		// Create new destination trie for this source prefix
+		destTriePtr = &destTrie{
+			trie:  &bart.Table[[]*SubnetRule]{},
+			rules: make([]*SubnetRule, 0),
+		}
+		sl.sourceTrie.Insert(canonicalSourcePrefix, destTriePtr)
+	}
+
+	// Canonicalize destination prefix to handle host bits correctly
+	// BART masks prefixes internally, so we need to match that behavior in our bookkeeping
+	canonicalDestPrefix := destPrefix.Masked()
+
+	// Add rule to destination trie
+	// Original behavior: overwrite if same (sourcePrefix, destPrefix) exists
+	// Store as single-element slice to match original overwrite behavior
+	destTriePtr.trie.Insert(canonicalDestPrefix, []*SubnetRule{rule})
+
+	// Update destTriePtr.rules - remove old rule with same canonical prefix if exists, then add new one
+	// Use canonical comparison to handle cases like 10.0.0.5/24 vs 10.0.0.0/24
+	newRules := make([]*SubnetRule, 0, len(destTriePtr.rules)+1)
+	for _, r := range destTriePtr.rules {
+		if !prefixEqual(r.DestPrefix, canonicalDestPrefix) || !prefixEqual(r.SourcePrefix, canonicalSourcePrefix) {
+			newRules = append(newRules, r)
+		}
+	}
+	newRules = append(newRules, rule)
+	destTriePtr.rules = newRules
+}
+
+// RemoveSubnet removes a subnet rule from the lookup table
+func (sl *SubnetLookup) RemoveSubnet(sourcePrefix, destPrefix netip.Prefix) {
+	sl.mu.Lock()
+	defer sl.mu.Unlock()
+
+	// Canonicalize prefixes to handle host bits correctly
+	canonicalSourcePrefix := sourcePrefix.Masked()
+	canonicalDestPrefix := destPrefix.Masked()
+
+	destTriePtr, exists := sl.sourceTrie.Get(canonicalSourcePrefix)
+	if !exists {
+		return
+	}
+
+	// Remove the rule - original behavior: delete exact (sourcePrefix, destPrefix) combination
+	// BART masks prefixes internally, so Delete works with canonical form
+	destTriePtr.trie.Delete(canonicalDestPrefix)
+
+	// Also remove from destTriePtr.rules using canonical comparison
+	// This ensures we remove rules even if they were added with host bits set
+	newDestRules := make([]*SubnetRule, 0, len(destTriePtr.rules))
+	for _, r := range destTriePtr.rules {
+		if !prefixEqual(r.DestPrefix, canonicalDestPrefix) || !prefixEqual(r.SourcePrefix, canonicalSourcePrefix) {
+			newDestRules = append(newDestRules, r)
+		}
+	}
+	destTriePtr.rules = newDestRules
+
+	// Check if the trie is actually empty using BART's Size() method
+	// This is more efficient than iterating and ensures we clean up empty tries
+	// even if there were stale entries in the rules slice (which shouldn't happen
+	// with proper canonicalization, but this provides a definitive check)
+	if destTriePtr.trie.Size() == 0 {
+		sl.sourceTrie.Delete(canonicalSourcePrefix)
+	}
+}
+
+// Match checks if a source IP, destination IP, port, and protocol match any subnet rule
+// Returns the matched rule if ALL of these conditions are met:
+//   - The source IP is in the rule's source prefix
+//   - The destination IP is in the rule's destination prefix
+//   - The port is in an allowed range (or no port restrictions exist)
+//   - The protocol matches (or the port range allows both protocols)
+//
+// proto should be header.TCPProtocolNumber, header.UDPProtocolNumber, or header.ICMPv4ProtocolNumber
+// Returns nil if no rule matches
+// This uses BART's Supernets() for O(log n) prefix matching instead of O(n) iteration
+func (sl *SubnetLookup) Match(srcIP, dstIP netip.Addr, port uint16, proto tcpip.TransportProtocolNumber) *SubnetRule {
+	sl.mu.RLock()
+	defer sl.mu.RUnlock()
+
+	// Convert IP addresses to /32 (IPv4) or /128 (IPv6) prefixes
+	// Supernets() finds all prefixes that contain this IP (i.e., are supernets of /32 or /128)
+	srcPrefix := netip.PrefixFrom(srcIP, srcIP.BitLen())
+	dstPrefix := netip.PrefixFrom(dstIP, dstIP.BitLen())
+
+	// Step 1: Find all source prefixes that contain srcIP using BART's Supernets
+	// This is O(log n) instead of O(n) iteration
+	// Supernets returns all prefixes that are supernets (contain) the given prefix
+	for _, destTriePtr := range sl.sourceTrie.Supernets(srcPrefix) {
+		if destTriePtr == nil {
+			continue
+		}
+
+		// Step 2: Find all destination prefixes that contain dstIP
+		// This is also O(log n) for each matching source prefix
+		for _, rules := range destTriePtr.trie.Supernets(dstPrefix) {
+			if rules == nil {
+				continue
+			}
+
+			// Step 3: Check each rule for ICMP and port restrictions
+			for _, rule := range rules {
+				// Handle ICMP before port range check — ICMP has no ports
+				if proto == header.ICMPv4ProtocolNumber || proto == header.ICMPv6ProtocolNumber {
+					if rule.DisableIcmp {
+						return nil
+					}
+					// ICMP is allowed; port ranges don't apply to ICMP
+					return rule
+				}
+
+				// Check port restrictions
+				if len(rule.PortRanges) == 0 {
+					// No port restrictions, match!
+					return rule
+				}
+
+				// Check if port and protocol are in any of the allowed ranges
+				for _, pr := range rule.PortRanges {
+					if port >= pr.Min && port <= pr.Max {
+						// Check protocol compatibility
+						if pr.Protocol == "" {
+							// Empty protocol means allow both TCP and UDP
+							return rule
+						}
+						// Check if the packet protocol matches the port range protocol
+						if (pr.Protocol == "tcp" && proto == header.TCPProtocolNumber) ||
+							(pr.Protocol == "udp" && proto == header.UDPProtocolNumber) {
+							return rule
+						}
+						// Port matches but protocol doesn't - continue checking other ranges
+					}
+				}
+			}
+		}
+	}
+
+	return nil
+}

netstack2/tun.go

@@ -56,15 +56,17 @@ type Net netTun
 
 // NetTunOptions contains options for creating a NetTUN device
 type NetTunOptions struct {
-	EnableTCPProxy bool
-	EnableUDPProxy bool
+	EnableTCPProxy  bool
+	EnableUDPProxy  bool
+	EnableICMPProxy bool
 }
 
 // CreateNetTUN creates a new TUN device with netstack without proxying
 func CreateNetTUN(localAddresses, dnsServers []netip.Addr, mtu int) (tun.Device, *Net, error) {
 	return CreateNetTUNWithOptions(localAddresses, dnsServers, mtu, NetTunOptions{
-		EnableTCPProxy: true,
-		EnableUDPProxy: true,
+		EnableTCPProxy:  true,
+		EnableUDPProxy:  true,
+		EnableICMPProxy: true,
 	})
 }
 
@@ -84,13 +86,14 @@ func CreateNetTUNWithOptions(localAddresses, dnsServers []netip.Addr, mtu int, o
 		mtu:            mtu,
 	}
 
-	// Initialize proxy handler if TCP or UDP proxying is enabled
-	if options.EnableTCPProxy || options.EnableUDPProxy {
+	// Initialize proxy handler if TCP, UDP, or ICMP proxying is enabled
+	if options.EnableTCPProxy || options.EnableUDPProxy || options.EnableICMPProxy {
 		var err error
 		dev.proxyHandler, err = NewProxyHandler(ProxyHandlerOptions{
-			EnableTCP: options.EnableTCPProxy,
-			EnableUDP: options.EnableUDPProxy,
-			MTU:       mtu,
+			EnableTCP:  options.EnableTCPProxy,
+			EnableUDP:  options.EnableUDPProxy,
+			EnableICMP: options.EnableICMPProxy,
+			MTU:        mtu,
 		})
 		if err != nil {
 			return nil, nil, fmt.Errorf("failed to create proxy handler: %v", err)
@@ -351,10 +354,10 @@ func (net *Net) ListenUDP(laddr *net.UDPAddr) (*gonet.UDPConn, error) {
 // AddProxySubnetRule adds a subnet rule to the proxy handler
 // If portRanges is nil or empty, all ports are allowed for this subnet
 // rewriteTo can be either an IP/CIDR (e.g., "192.168.1.1/32") or a domain name (e.g., "example.com")
-func (net *Net) AddProxySubnetRule(sourcePrefix, destPrefix netip.Prefix, rewriteTo string, portRanges []PortRange) {
+func (net *Net) AddProxySubnetRule(sourcePrefix, destPrefix netip.Prefix, rewriteTo string, portRanges []PortRange, disableIcmp bool) {
 	tun := (*netTun)(net)
 	if tun.proxyHandler != nil {
-		tun.proxyHandler.AddSubnetRule(sourcePrefix, destPrefix, rewriteTo, portRanges)
+		tun.proxyHandler.AddSubnetRule(sourcePrefix, destPrefix, rewriteTo, portRanges, disableIcmp)
 	}
 }
 

network/interface.go

@@ -44,9 +44,13 @@ func ConfigureInterface(interfaceName string, tunnelIp string, mtu int) error {
 		return configureDarwin(interfaceName, ip, ipNet)
 	case "windows":
 		return configureWindows(interfaceName, ip, ipNet)
-	default:
-		return fmt.Errorf("unsupported operating system: %s", runtime.GOOS)
+	case "android":
+		return nil
+	case "ios":
+		return nil
 	}
+
+	return nil
 }
 
 // waitForInterfaceUp polls the network interface until it's up or times out

network/route.go

@@ -126,13 +126,14 @@ func LinuxRemoveRoute(destination string) error {
 
 // addRouteForServerIP adds an OS-specific route for the server IP
 func AddRouteForServerIP(serverIP, interfaceName string) error {
-	if err := AddRouteForNetworkConfig(serverIP); err != nil {
-		return err
-	}
 	if interfaceName == "" {
 		return nil
 	}
-	if runtime.GOOS == "darwin" {
+	// TODO: does this also need to be ios?
+	if runtime.GOOS == "darwin" { // macos requires routes for each peer to be added but this messes with other platforms
+		if err := AddRouteForNetworkConfig(serverIP); err != nil {
+			return err
+		}
 		return DarwinAddRoute(serverIP, "", interfaceName)
 	}
 	// else if runtime.GOOS == "windows" {
@@ -145,13 +146,14 @@ func AddRouteForServerIP(serverIP, interfaceName string) error {
 
 // removeRouteForServerIP removes an OS-specific route for the server IP
 func RemoveRouteForServerIP(serverIP string, interfaceName string) error {
-	if err := RemoveRouteForNetworkConfig(serverIP); err != nil {
-		return err
-	}
 	if interfaceName == "" {
 		return nil
 	}
-	if runtime.GOOS == "darwin" {
+	// TODO: does this also need to be ios?
+	if runtime.GOOS == "darwin" { // macos requires routes for each peer to be added but this messes with other platforms
+		if err := RemoveRouteForNetworkConfig(serverIP); err != nil {
+			return err
+		}
 		return DarwinRemoveRoute(serverIP)
 	}
 	// else if runtime.GOOS == "windows" {
@@ -217,21 +219,22 @@ func AddRoutes(remoteSubnets []string, interfaceName string) error {
 			continue
 		}
 
-		if runtime.GOOS == "darwin" {
+		switch runtime.GOOS {
+		case "darwin":
 			if err := DarwinAddRoute(subnet, "", interfaceName); err != nil {
 				logger.Error("Failed to add Darwin route for subnet %s: %v", subnet, err)
-				return err
 			}
-		} else if runtime.GOOS == "windows" {
+		case "windows":
 			if err := WindowsAddRoute(subnet, "", interfaceName); err != nil {
 				logger.Error("Failed to add Windows route for subnet %s: %v", subnet, err)
-				return err
 			}
-		} else if runtime.GOOS == "linux" {
+		case "linux":
 			if err := LinuxAddRoute(subnet, "", interfaceName); err != nil {
 				logger.Error("Failed to add Linux route for subnet %s: %v", subnet, err)
-				return err
 			}
+		case "android", "ios":
+			// Routes handled by the OS/VPN service
+			continue
 		}
 
 		logger.Info("Added route for remote subnet: %s", subnet)
@@ -258,21 +261,22 @@ func RemoveRoutes(remoteSubnets []string) error {
 		}
 
 		// Remove route based on operating system
-		if runtime.GOOS == "darwin" {
+		switch runtime.GOOS {
+		case "darwin":
 			if err := DarwinRemoveRoute(subnet); err != nil {
 				logger.Error("Failed to remove Darwin route for subnet %s: %v", subnet, err)
-				return err
 			}
-		} else if runtime.GOOS == "windows" {
+		case "windows":
 			if err := WindowsRemoveRoute(subnet); err != nil {
 				logger.Error("Failed to remove Windows route for subnet %s: %v", subnet, err)
-				return err
 			}
-		} else if runtime.GOOS == "linux" {
+		case "linux":
 			if err := LinuxRemoveRoute(subnet); err != nil {
 				logger.Error("Failed to remove Linux route for subnet %s: %v", subnet, err)
-				return err
 			}
+		case "android", "ios":
+			// Routes handled by the OS/VPN service
+			continue
 		}
 
 		logger.Info("Removed route for remote subnet: %s", subnet)

network/settings.go

@@ -115,7 +115,7 @@ func RemoveIPv4IncludedRoute(route IPv4Route) {
 		if r == route {
 			networkSettings.IPv4IncludedRoutes = append(routes[:i], routes[i+1:]...)
 			logger.Info("Removed IPv4 included route: %+v", route)
-			return
+			break
 		}
 	}
 	incrementor++

newt.iss

@@ -0,0 +1,152 @@
+; Script generated by the Inno Setup Script Wizard.
+; SEE THE DOCUMENTATION FOR DETAILS ON CREATING INNO SETUP SCRIPT FILES!
+
+#define MyAppName "newt"
+#define MyAppVersion "1.0.0"
+#define MyAppPublisher "Fossorial Inc."
+#define MyAppURL "https://pangolin.net"
+#define MyAppExeName "newt.exe"
+
+[Setup]
+; NOTE: The value of AppId uniquely identifies this application. Do not use the same AppId value in installers for other applications.
+; (To generate a new GUID, click Tools | Generate GUID inside the IDE.)
+AppId={{25A1E3C4-F273-4334-8DF3-47408E83012D}
+AppName={#MyAppName}
+AppVersion={#MyAppVersion}
+;AppVerName={#MyAppName} {#MyAppVersion}
+AppPublisher={#MyAppPublisher}
+AppPublisherURL={#MyAppURL}
+AppSupportURL={#MyAppURL}
+AppUpdatesURL={#MyAppURL}
+DefaultDirName={autopf}\{#MyAppName}
+UninstallDisplayIcon={app}\{#MyAppExeName}
+; "ArchitecturesAllowed=x64compatible" specifies that Setup cannot run
+; on anything but x64 and Windows 11 on Arm.
+ArchitecturesAllowed=x64compatible
+; "ArchitecturesInstallIn64BitMode=x64compatible" requests that the
+; install be done in "64-bit mode" on x64 or Windows 11 on Arm,
+; meaning it should use the native 64-bit Program Files directory and
+; the 64-bit view of the registry.
+ArchitecturesInstallIn64BitMode=x64compatible
+DefaultGroupName={#MyAppName}
+DisableProgramGroupPage=yes
+; Uncomment the following line to run in non administrative install mode (install for current user only).
+;PrivilegesRequired=lowest
+OutputBaseFilename=newt_windows_installer
+SolidCompression=yes
+WizardStyle=modern
+; Add this to ensure PATH changes are applied and the system is prompted for a restart if needed
+RestartIfNeededByRun=no
+ChangesEnvironment=true
+
+[Languages]
+Name: "english"; MessagesFile: "compiler:Default.isl"
+
+[Files]
+; The 'DestName' flag ensures that 'newt_windows_amd64.exe' is installed as 'newt.exe'
+Source: "Z:\newt_windows_amd64.exe"; DestDir: "{app}"; DestName: "{#MyAppExeName}"; Flags: ignoreversion
+Source: "Z:\wintun.dll"; DestDir: "{app}"; Flags: ignoreversion
+; NOTE: Don't use "Flags: ignoreversion" on any shared system files
+
+[Icons]
+Name: "{group}\{#MyAppName}"; Filename: "{app}\{#MyAppExeName}"
+
+[Registry]
+; Add the application's installation directory to the system PATH environment variable.
+; HKLM (HKEY_LOCAL_MACHINE) is used for system-wide changes.
+; The 'Path' variable is located under 'SYSTEM\CurrentControlSet\Control\Session Manager\Environment'.
+; ValueType: expandsz allows for environment variables (like %ProgramFiles%) in the path.
+; ValueData: "{olddata};{app}" appends the current application directory to the existing PATH.
+; Note: Removal during uninstallation is handled by CurUninstallStepChanged procedure in [Code] section.
+; Check: NeedsAddPath ensures this is applied only if the path is not already present.
+[Registry]
+; Add the application's installation directory to the system PATH.
+Root: HKLM; Subkey: "SYSTEM\CurrentControlSet\Control\Session Manager\Environment"; \
+    ValueType: expandsz; ValueName: "Path"; ValueData: "{olddata};{app}"; \
+    Check: NeedsAddPath(ExpandConstant('{app}'))
+
+[Code]
+function NeedsAddPath(Path: string): boolean;
+var
+  OrigPath: string;
+begin
+  if not RegQueryStringValue(HKEY_LOCAL_MACHINE,
+    'SYSTEM\CurrentControlSet\Control\Session Manager\Environment',
+    'Path', OrigPath)
+  then begin
+    // Path variable doesn't exist at all, so we definitely need to add it.
+    Result := True;
+    exit;
+  end;
+
+  // Perform a case-insensitive check to see if the path is already present.
+  // We add semicolons to prevent partial matches (e.g., matching C:\App in C:\App2).
+  if Pos(';' + UpperCase(Path) + ';', ';' + UpperCase(OrigPath) + ';') > 0 then
+    Result := False
+  else
+    Result := True;
+end;
+
+procedure RemovePathEntry(PathToRemove: string);
+var
+  OrigPath: string;
+  NewPath: string;
+  PathList: TStringList;
+  I: Integer;
+begin
+  if not RegQueryStringValue(HKEY_LOCAL_MACHINE,
+    'SYSTEM\CurrentControlSet\Control\Session Manager\Environment',
+    'Path', OrigPath)
+  then begin
+    // Path variable doesn't exist, nothing to remove
+    exit;
+  end;
+
+  // Create a string list to parse the PATH entries
+  PathList := TStringList.Create;
+  try
+    // Split the PATH by semicolons
+    PathList.Delimiter := ';';
+    PathList.StrictDelimiter := True;
+    PathList.DelimitedText := OrigPath;
+
+    // Find and remove the matching entry (case-insensitive)
+    for I := PathList.Count - 1 downto 0 do
+    begin
+      if CompareText(Trim(PathList[I]), Trim(PathToRemove)) = 0 then
+      begin
+        Log('Found and removing PATH entry: ' + PathList[I]);
+        PathList.Delete(I);
+      end;
+    end;
+
+    // Reconstruct the PATH
+    NewPath := PathList.DelimitedText;
+
+    // Write the new PATH back to the registry
+    if RegWriteExpandStringValue(HKEY_LOCAL_MACHINE,
+      'SYSTEM\CurrentControlSet\Control\Session Manager\Environment',
+      'Path', NewPath)
+    then
+      Log('Successfully removed path entry: ' + PathToRemove)
+    else
+      Log('Failed to write modified PATH to registry');
+  finally
+    PathList.Free;
+  end;
+end;
+
+procedure CurUninstallStepChanged(CurUninstallStep: TUninstallStep);
+var
+  AppPath: string;
+begin
+  if CurUninstallStep = usUninstall then
+  begin
+    // Get the application installation path
+    AppPath := ExpandConstant('{app}');
+    Log('Removing PATH entry for: ' + AppPath);
+
+    // Remove only our path entry from the system PATH
+    RemovePathEntry(AppPath);
+  end;
+end;

service_unix.go

@@ -0,0 +1,59 @@
+//go:build !windows
+
+package main
+
+import (
+	"fmt"
+)
+
+// Service management functions are not available on non-Windows platforms
+func installService() error {
+	return fmt.Errorf("service management is only available on Windows")
+}
+
+func removeService() error {
+	return fmt.Errorf("service management is only available on Windows")
+}
+
+func startService(args []string) error {
+	_ = args // unused on Unix platforms
+	return fmt.Errorf("service management is only available on Windows")
+}
+
+func stopService() error {
+	return fmt.Errorf("service management is only available on Windows")
+}
+
+func getServiceStatus() (string, error) {
+	return "", fmt.Errorf("service management is only available on Windows")
+}
+
+func debugService(args []string) error {
+	_ = args // unused on Unix platforms
+	return fmt.Errorf("debug service is only available on Windows")
+}
+
+func isWindowsService() bool {
+	return false
+}
+
+func runService(name string, isDebug bool, args []string) {
+	// No-op on non-Windows platforms
+}
+
+func setupWindowsEventLog() {
+	// No-op on non-Windows platforms
+}
+
+func watchLogFile(end bool) error {
+	return fmt.Errorf("watching log file is only available on Windows")
+}
+
+func showServiceConfig() {
+	fmt.Println("Service configuration is only available on Windows")
+}
+
+// handleServiceCommand returns false on non-Windows platforms
+func handleServiceCommand() bool {
+	return false
+}

service_windows.go

@@ -0,0 +1,760 @@
+//go:build windows
+
+package main
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"log"
+	"os"
+	"os/signal"
+	"path/filepath"
+	"strings"
+	"syscall"
+	"time"
+
+	"github.com/fosrl/newt/logger"
+	"golang.org/x/sys/windows/svc"
+	"golang.org/x/sys/windows/svc/debug"
+	"golang.org/x/sys/windows/svc/eventlog"
+	"golang.org/x/sys/windows/svc/mgr"
+)
+
+const (
+	serviceName        = "NewtWireguardService"
+	serviceDisplayName = "Newt WireGuard Tunnel Service"
+	serviceDescription = "Newt WireGuard tunnel service for secure network connectivity"
+)
+
+// Global variable to store service arguments
+var serviceArgs []string
+
+// getServiceArgsPath returns the path where service arguments are stored
+func getServiceArgsPath() string {
+	logDir := filepath.Join(os.Getenv("PROGRAMDATA"), "newt")
+	return filepath.Join(logDir, "service_args.json")
+}
+
+// saveServiceArgs saves the service arguments to a file
+func saveServiceArgs(args []string) error {
+	logDir := filepath.Join(os.Getenv("PROGRAMDATA"), "newt")
+	err := os.MkdirAll(logDir, 0755)
+	if err != nil {
+		return fmt.Errorf("failed to create config directory: %v", err)
+	}
+
+	argsPath := getServiceArgsPath()
+	data, err := json.Marshal(args)
+	if err != nil {
+		return fmt.Errorf("failed to marshal service args: %v", err)
+	}
+
+	err = os.WriteFile(argsPath, data, 0644)
+	if err != nil {
+		return fmt.Errorf("failed to write service args: %v", err)
+	}
+
+	return nil
+}
+
+// loadServiceArgs loads the service arguments from a file
+func loadServiceArgs() ([]string, error) {
+	argsPath := getServiceArgsPath()
+	data, err := os.ReadFile(argsPath)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return []string{}, nil // Return empty args if file doesn't exist
+		}
+		return nil, fmt.Errorf("failed to read service args: %v", err)
+	}
+
+	var args []string
+	err = json.Unmarshal(data, &args)
+	if err != nil {
+		return nil, fmt.Errorf("failed to unmarshal service args: %v", err)
+	}
+
+	return args, nil
+}
+
+type newtService struct {
+	elog debug.Log
+	ctx  context.Context
+	stop context.CancelFunc
+	args []string
+}
+
+func (s *newtService) Execute(args []string, r <-chan svc.ChangeRequest, changes chan<- svc.Status) (bool, uint32) {
+	const cmdsAccepted = svc.AcceptStop | svc.AcceptShutdown
+	changes <- svc.Status{State: svc.StartPending}
+
+	s.elog.Info(1, fmt.Sprintf("Service Execute called with args: %v", args))
+
+	// Load saved service arguments
+	savedArgs, err := loadServiceArgs()
+	if err != nil {
+		s.elog.Error(1, fmt.Sprintf("Failed to load service args: %v", err))
+		// Continue with empty args if loading fails
+		savedArgs = []string{}
+	}
+	s.elog.Info(1, fmt.Sprintf("Loaded saved service args: %v", savedArgs))
+
+	// Combine service start args with saved args, giving priority to service start args
+	// Note: When the service is started via SCM, args[0] is the service name
+	// When started via s.Start(args...), the args passed are exactly what we provide
+	finalArgs := []string{}
+
+	// Check if we have args passed directly to Execute (from s.Start())
+	if len(args) > 0 {
+		// The first arg from SCM is the service name, but when we call s.Start(args...),
+		// the args we pass become args[1:] in Execute. However, if started by SCM without
+		// args, args[0] will be the service name.
+		// We need to check if args[0] looks like the service name or a flag
+		if len(args) == 1 && args[0] == serviceName {
+			// Only service name, no actual args
+			s.elog.Info(1, "Only service name in args, checking saved args")
+		} else if len(args) > 1 && args[0] == serviceName {
+			// Service name followed by actual args
+			finalArgs = append(finalArgs, args[1:]...)
+			s.elog.Info(1, fmt.Sprintf("Using service start parameters (after service name): %v", finalArgs))
+		} else {
+			// Args don't start with service name, use them all
+			// This happens when args are passed via s.Start(args...)
+			finalArgs = append(finalArgs, args...)
+			s.elog.Info(1, fmt.Sprintf("Using service start parameters (direct): %v", finalArgs))
+		}
+	}
+
+	// If no service start parameters, use saved args
+	if len(finalArgs) == 0 && len(savedArgs) > 0 {
+		finalArgs = savedArgs
+		s.elog.Info(1, fmt.Sprintf("Using saved service args: %v", finalArgs))
+	}
+
+	s.elog.Info(1, fmt.Sprintf("Final args to use: %v", finalArgs))
+	s.args = finalArgs
+
+	// Start the main newt functionality
+	newtDone := make(chan struct{})
+	go func() {
+		s.runNewt()
+		close(newtDone)
+	}()
+
+	changes <- svc.Status{State: svc.Running, Accepts: cmdsAccepted}
+	s.elog.Info(1, "Service status set to Running")
+
+	for {
+		select {
+		case c := <-r:
+			switch c.Cmd {
+			case svc.Interrogate:
+				changes <- c.CurrentStatus
+			case svc.Stop, svc.Shutdown:
+				s.elog.Info(1, "Service stopping")
+				changes <- svc.Status{State: svc.StopPending}
+				if s.stop != nil {
+					s.stop()
+				}
+				// Wait for main logic to finish or timeout
+				select {
+				case <-newtDone:
+					s.elog.Info(1, "Main logic finished gracefully")
+				case <-time.After(10 * time.Second):
+					s.elog.Info(1, "Timeout waiting for main logic to finish")
+				}
+				return false, 0
+			default:
+				s.elog.Error(1, fmt.Sprintf("Unexpected control request #%d", c))
+			}
+		case <-newtDone:
+			s.elog.Info(1, "Main newt logic completed, stopping service")
+			changes <- svc.Status{State: svc.StopPending}
+			return false, 0
+		}
+	}
+}
+
+func (s *newtService) runNewt() {
+	// Create a context that can be cancelled when the service stops
+	s.ctx, s.stop = context.WithCancel(context.Background())
+
+	// Setup logging for service mode
+	s.elog.Info(1, "Starting Newt main logic")
+
+	// Run the main newt logic and wait for it to complete
+	done := make(chan struct{})
+	go func() {
+		defer func() {
+			if r := recover(); r != nil {
+				s.elog.Error(1, fmt.Sprintf("Panic in newt main: %v", r))
+			}
+			close(done)
+		}()
+
+		// Call the main newt function with stored arguments
+		// Use s.ctx as the signal context since the service manages shutdown
+		runNewtMainWithArgs(s.ctx, s.args)
+	}()
+
+	// Wait for either context cancellation or main logic completion
+	select {
+	case <-s.ctx.Done():
+		s.elog.Info(1, "Newt service context cancelled")
+	case <-done:
+		s.elog.Info(1, "Newt main logic completed")
+	}
+}
+
+func runService(name string, isDebug bool, args []string) {
+	var err error
+	var elog debug.Log
+
+	if isDebug {
+		elog = debug.New(name)
+		fmt.Printf("Starting %s service in debug mode\n", name)
+	} else {
+		elog, err = eventlog.Open(name)
+		if err != nil {
+			fmt.Printf("Failed to open event log: %v\n", err)
+			return
+		}
+	}
+	defer elog.Close()
+
+	elog.Info(1, fmt.Sprintf("Starting %s service", name))
+	run := svc.Run
+	if isDebug {
+		run = debug.Run
+	}
+
+	service := &newtService{elog: elog, args: args}
+	err = run(name, service)
+	if err != nil {
+		elog.Error(1, fmt.Sprintf("%s service failed: %v", name, err))
+		if isDebug {
+			fmt.Printf("Service failed: %v\n", err)
+		}
+		return
+	}
+	elog.Info(1, fmt.Sprintf("%s service stopped", name))
+	if isDebug {
+		fmt.Printf("%s service stopped\n", name)
+	}
+}
+
+func installService() error {
+	exepath, err := os.Executable()
+	if err != nil {
+		return fmt.Errorf("failed to get executable path: %v", err)
+	}
+
+	m, err := mgr.Connect()
+	if err != nil {
+		return fmt.Errorf("failed to connect to service manager: %v", err)
+	}
+	defer m.Disconnect()
+
+	s, err := m.OpenService(serviceName)
+	if err == nil {
+		s.Close()
+		return fmt.Errorf("service %s already exists", serviceName)
+	}
+
+	config := mgr.Config{
+		ServiceType:    0x10, // SERVICE_WIN32_OWN_PROCESS
+		StartType:      mgr.StartManual,
+		ErrorControl:   mgr.ErrorNormal,
+		DisplayName:    serviceDisplayName,
+		Description:    serviceDescription,
+		BinaryPathName: exepath,
+	}
+
+	s, err = m.CreateService(serviceName, exepath, config)
+	if err != nil {
+		return fmt.Errorf("failed to create service: %v", err)
+	}
+	defer s.Close()
+
+	err = eventlog.InstallAsEventCreate(serviceName, eventlog.Error|eventlog.Warning|eventlog.Info)
+	if err != nil {
+		s.Delete()
+		return fmt.Errorf("failed to install event log: %v", err)
+	}
+
+	return nil
+}
+
+func removeService() error {
+	m, err := mgr.Connect()
+	if err != nil {
+		return fmt.Errorf("failed to connect to service manager: %v", err)
+	}
+	defer m.Disconnect()
+
+	s, err := m.OpenService(serviceName)
+	if err != nil {
+		return fmt.Errorf("service %s is not installed", serviceName)
+	}
+	defer s.Close()
+
+	// Stop the service if it's running
+	status, err := s.Query()
+	if err != nil {
+		return fmt.Errorf("failed to query service status: %v", err)
+	}
+
+	if status.State != svc.Stopped {
+		_, err = s.Control(svc.Stop)
+		if err != nil {
+			return fmt.Errorf("failed to stop service: %v", err)
+		}
+
+		// Wait for service to stop
+		timeout := time.Now().Add(30 * time.Second)
+		for status.State != svc.Stopped {
+			if timeout.Before(time.Now()) {
+				return fmt.Errorf("timeout waiting for service to stop")
+			}
+			time.Sleep(300 * time.Millisecond)
+			status, err = s.Query()
+			if err != nil {
+				return fmt.Errorf("failed to query service status: %v", err)
+			}
+		}
+	}
+
+	err = s.Delete()
+	if err != nil {
+		return fmt.Errorf("failed to delete service: %v", err)
+	}
+
+	err = eventlog.Remove(serviceName)
+	if err != nil {
+		return fmt.Errorf("failed to remove event log: %v", err)
+	}
+
+	return nil
+}
+
+func startService(args []string) error {
+	fmt.Printf("Starting service with args: %v\n", args)
+
+	// Always save the service arguments so they can be loaded on service restart
+	err := saveServiceArgs(args)
+	if err != nil {
+		fmt.Printf("Warning: failed to save service args: %v\n", err)
+		// Continue anyway, args will still be passed directly
+	} else {
+		fmt.Printf("Saved service args to: %s\n", getServiceArgsPath())
+	}
+
+	m, err := mgr.Connect()
+	if err != nil {
+		return fmt.Errorf("failed to connect to service manager: %v", err)
+	}
+	defer m.Disconnect()
+
+	s, err := m.OpenService(serviceName)
+	if err != nil {
+		return fmt.Errorf("service %s is not installed", serviceName)
+	}
+	defer s.Close()
+
+	// Pass arguments directly to the service start call
+	// Note: These args will appear in Execute() after the service name
+	err = s.Start(args...)
+	if err != nil {
+		return fmt.Errorf("failed to start service: %v", err)
+	}
+
+	return nil
+}
+
+func stopService() error {
+	m, err := mgr.Connect()
+	if err != nil {
+		return fmt.Errorf("failed to connect to service manager: %v", err)
+	}
+	defer m.Disconnect()
+
+	s, err := m.OpenService(serviceName)
+	if err != nil {
+		return fmt.Errorf("service %s is not installed", serviceName)
+	}
+	defer s.Close()
+
+	status, err := s.Control(svc.Stop)
+	if err != nil {
+		return fmt.Errorf("failed to stop service: %v", err)
+	}
+
+	timeout := time.Now().Add(30 * time.Second)
+	for status.State != svc.Stopped {
+		if timeout.Before(time.Now()) {
+			return fmt.Errorf("timeout waiting for service to stop")
+		}
+		time.Sleep(300 * time.Millisecond)
+		status, err = s.Query()
+		if err != nil {
+			return fmt.Errorf("failed to query service status: %v", err)
+		}
+	}
+
+	return nil
+}
+
+func debugService(args []string) error {
+	// Save the service arguments before starting
+	if len(args) > 0 {
+		err := saveServiceArgs(args)
+		if err != nil {
+			return fmt.Errorf("failed to save service args: %v", err)
+		}
+	}
+
+	// Start the service with the provided arguments
+	err := startService(args)
+	if err != nil {
+		return fmt.Errorf("failed to start service: %v", err)
+	}
+
+	// Watch the log file
+	return watchLogFile(true)
+}
+
+func watchLogFile(end bool) error {
+	logDir := filepath.Join(os.Getenv("PROGRAMDATA"), "newt", "logs")
+	logPath := filepath.Join(logDir, "newt.log")
+
+	// Ensure the log directory exists
+	err := os.MkdirAll(logDir, 0755)
+	if err != nil {
+		return fmt.Errorf("failed to create log directory: %v", err)
+	}
+
+	// Wait for the log file to be created if it doesn't exist
+	var file *os.File
+	for i := 0; i < 30; i++ { // Wait up to 15 seconds
+		file, err = os.Open(logPath)
+		if err == nil {
+			break
+		}
+		if i == 0 {
+			fmt.Printf("Waiting for log file to be created...\n")
+		}
+		time.Sleep(500 * time.Millisecond)
+	}
+
+	if err != nil {
+		return fmt.Errorf("failed to open log file after waiting: %v", err)
+	}
+	defer file.Close()
+
+	// Seek to the end of the file to only show new logs
+	_, err = file.Seek(0, 2)
+	if err != nil {
+		return fmt.Errorf("failed to seek to end of file: %v", err)
+	}
+
+	// Set up signal handling for graceful exit
+	sigCh := make(chan os.Signal, 1)
+	signal.Notify(sigCh, os.Interrupt, syscall.SIGTERM)
+
+	// Create a ticker to check for new content
+	ticker := time.NewTicker(500 * time.Millisecond)
+	defer ticker.Stop()
+
+	buffer := make([]byte, 4096)
+
+	for {
+		select {
+		case <-sigCh:
+			fmt.Printf("\n\nStopping log watch...\n")
+			// stop the service if needed
+			if end {
+				fmt.Printf("Stopping service...\n")
+				stopService()
+			}
+			fmt.Printf("Log watch stopped.\n")
+			return nil
+		case <-ticker.C:
+			// Read new content
+			n, err := file.Read(buffer)
+			if err != nil && err != io.EOF {
+				// Try to reopen the file in case it was recreated
+				file.Close()
+				file, err = os.Open(logPath)
+				if err != nil {
+					continue
+				}
+				continue
+			}
+
+			if n > 0 {
+				// Print the new content
+				fmt.Print(string(buffer[:n]))
+			}
+		}
+	}
+}
+
+func getServiceStatus() (string, error) {
+	m, err := mgr.Connect()
+	if err != nil {
+		return "", fmt.Errorf("failed to connect to service manager: %v", err)
+	}
+	defer m.Disconnect()
+
+	s, err := m.OpenService(serviceName)
+	if err != nil {
+		return "Not Installed", nil
+	}
+	defer s.Close()
+
+	status, err := s.Query()
+	if err != nil {
+		return "", fmt.Errorf("failed to query service status: %v", err)
+	}
+
+	switch status.State {
+	case svc.Stopped:
+		return "Stopped", nil
+	case svc.StartPending:
+		return "Starting", nil
+	case svc.StopPending:
+		return "Stopping", nil
+	case svc.Running:
+		return "Running", nil
+	case svc.ContinuePending:
+		return "Continue Pending", nil
+	case svc.PausePending:
+		return "Pause Pending", nil
+	case svc.Paused:
+		return "Paused", nil
+	default:
+		return "Unknown", nil
+	}
+}
+
+// showServiceConfig displays current saved service configuration
+func showServiceConfig() {
+	configPath := getServiceArgsPath()
+	fmt.Printf("Service configuration file: %s\n", configPath)
+
+	args, err := loadServiceArgs()
+	if err != nil {
+		fmt.Printf("No saved configuration found or error loading: %v\n", err)
+		return
+	}
+
+	if len(args) == 0 {
+		fmt.Println("No saved service arguments found")
+	} else {
+		fmt.Printf("Saved service arguments: %v\n", args)
+	}
+}
+
+func isWindowsService() bool {
+	isWindowsService, err := svc.IsWindowsService()
+	return err == nil && isWindowsService
+}
+
+// rotateLogFile handles daily log rotation
+func rotateLogFile(logDir string, logFile string) error {
+	// Get current log file info
+	info, err := os.Stat(logFile)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return nil // No current log file to rotate
+		}
+		return fmt.Errorf("failed to stat log file: %v", err)
+	}
+
+	// Check if log file is from today
+	now := time.Now()
+	fileTime := info.ModTime()
+
+	// If the log file is from today, no rotation needed
+	if now.Year() == fileTime.Year() && now.YearDay() == fileTime.YearDay() {
+		return nil
+	}
+
+	// Create rotated filename with date
+	rotatedName := fmt.Sprintf("newt-%s.log", fileTime.Format("2006-01-02"))
+	rotatedPath := filepath.Join(logDir, rotatedName)
+
+	// Rename current log file to dated filename
+	err = os.Rename(logFile, rotatedPath)
+	if err != nil {
+		return fmt.Errorf("failed to rotate log file: %v", err)
+	}
+
+	// Clean up old log files (keep last 30 days)
+	cleanupOldLogFiles(logDir, 30)
+
+	return nil
+}
+
+// cleanupOldLogFiles removes log files older than specified days
+func cleanupOldLogFiles(logDir string, daysToKeep int) {
+	cutoff := time.Now().AddDate(0, 0, -daysToKeep)
+
+	files, err := os.ReadDir(logDir)
+	if err != nil {
+		return
+	}
+
+	for _, file := range files {
+		if !file.IsDir() && strings.HasPrefix(file.Name(), "newt-") && strings.HasSuffix(file.Name(), ".log") {
+			filePath := filepath.Join(logDir, file.Name())
+			info, err := file.Info()
+			if err != nil {
+				continue
+			}
+
+			if info.ModTime().Before(cutoff) {
+				os.Remove(filePath)
+			}
+		}
+	}
+}
+
+func setupWindowsEventLog() {
+	// Create log directory if it doesn't exist
+	logDir := filepath.Join(os.Getenv("PROGRAMDATA"), "newt", "logs")
+	err := os.MkdirAll(logDir, 0755)
+	if err != nil {
+		fmt.Printf("Failed to create log directory: %v\n", err)
+		return
+	}
+
+	logFile := filepath.Join(logDir, "newt.log")
+
+	// Rotate log file if needed
+	err = rotateLogFile(logDir, logFile)
+	if err != nil {
+		fmt.Printf("Failed to rotate log file: %v\n", err)
+		// Continue anyway to create new log file
+	}
+
+	file, err := os.OpenFile(logFile, os.O_CREATE|os.O_WRONLY|os.O_APPEND, 0666)
+	if err != nil {
+		fmt.Printf("Failed to open log file: %v\n", err)
+		return
+	}
+
+	// Set the custom logger output
+	logger.GetLogger().SetOutput(file)
+
+	log.Printf("Newt service logging initialized - log file: %s", logFile)
+}
+
+// handleServiceCommand checks for service management commands and returns true if handled
+func handleServiceCommand() bool {
+	if len(os.Args) < 2 {
+		return false
+	}
+
+	command := os.Args[1]
+
+	switch command {
+	case "install":
+		err := installService()
+		if err != nil {
+			fmt.Printf("Failed to install service: %v\n", err)
+			os.Exit(1)
+		}
+		fmt.Println("Service installed successfully")
+		return true
+	case "remove", "uninstall":
+		err := removeService()
+		if err != nil {
+			fmt.Printf("Failed to remove service: %v\n", err)
+			os.Exit(1)
+		}
+		fmt.Println("Service removed successfully")
+		return true
+	case "start":
+		// Pass the remaining arguments (after "start") to the service
+		serviceArgs := os.Args[2:]
+		err := startService(serviceArgs)
+		if err != nil {
+			fmt.Printf("Failed to start service: %v\n", err)
+			os.Exit(1)
+		}
+		fmt.Println("Service started successfully")
+		return true
+	case "stop":
+		err := stopService()
+		if err != nil {
+			fmt.Printf("Failed to stop service: %v\n", err)
+			os.Exit(1)
+		}
+		fmt.Println("Service stopped successfully")
+		return true
+	case "status":
+		status, err := getServiceStatus()
+		if err != nil {
+			fmt.Printf("Failed to get service status: %v\n", err)
+			os.Exit(1)
+		}
+		fmt.Printf("Service status: %s\n", status)
+		return true
+	case "debug":
+		// get the status and if it is Not Installed then install it first
+		status, err := getServiceStatus()
+		if err != nil {
+			fmt.Printf("Failed to get service status: %v\n", err)
+			os.Exit(1)
+		}
+		if status == "Not Installed" {
+			err := installService()
+			if err != nil {
+				fmt.Printf("Failed to install service: %v\n", err)
+				os.Exit(1)
+			}
+			fmt.Println("Service installed successfully, now running in debug mode")
+		}
+
+		// Pass the remaining arguments (after "debug") to the service
+		serviceArgs := os.Args[2:]
+		err = debugService(serviceArgs)
+		if err != nil {
+			fmt.Printf("Failed to debug service: %v\n", err)
+			os.Exit(1)
+		}
+		return true
+	case "logs":
+		err := watchLogFile(false)
+		if err != nil {
+			fmt.Printf("Failed to watch log file: %v\n", err)
+			os.Exit(1)
+		}
+		return true
+	case "config":
+		showServiceConfig()
+		return true
+	case "service-help":
+		fmt.Println("Newt WireGuard Tunnel")
+		fmt.Println("\nWindows Service Management:")
+		fmt.Println("  install        Install the service")
+		fmt.Println("  remove         Remove the service")
+		fmt.Println("  start [args]   Start the service with optional arguments")
+		fmt.Println("  stop           Stop the service")
+		fmt.Println("  status         Show service status")
+		fmt.Println("  debug [args]   Run service in debug mode with optional arguments")
+		fmt.Println("  logs           Tail the service log file")
+		fmt.Println("  config         Show current service configuration")
+		fmt.Println("  service-help   Show this service help")
+		fmt.Println("\nExamples:")
+		fmt.Println("  newt start --endpoint https://example.com --id myid --secret mysecret")
+		fmt.Println("  newt debug --endpoint https://example.com --id myid --secret mysecret")
+		fmt.Println("\nFor normal console mode, run with standard flags (e.g., newt --endpoint ...)")
+		return true
+	}
+
+	return false
+}

udp_client.py

@@ -0,0 +1,49 @@
+import socket
+import sys
+
+# Argument parsing: Check if IP and Port are provided
+if len(sys.argv) != 3:
+    print("Usage: python udp_client.py <HOST_IP> <HOST_PORT>")
+    # Example: python udp_client.py 127.0.0.1 12000
+    sys.exit(1)
+
+HOST = sys.argv[1]
+try:
+    PORT = int(sys.argv[2])
+except ValueError:
+    print("Error: HOST_PORT must be an integer.")
+    sys.exit(1)
+
+# The message to send to the server
+MESSAGE = "Hello UDP Server! How are you?"
+
+# Create a UDP socket
+try:
+    client_socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
+except socket.error as err:
+    print(f"Failed to create socket: {err}")
+    sys.exit()
+
+try:
+    print(f"Sending message to {HOST}:{PORT}...")
+
+    # Send the message (data must be encoded to bytes)
+    client_socket.sendto(MESSAGE.encode('utf-8'), (HOST, PORT))
+
+    # Wait for the server's response (buffer size 1024 bytes)
+    data, server_address = client_socket.recvfrom(1024)
+
+    # Decode and print the server's response
+    response = data.decode('utf-8')
+    print("-" * 30)
+    print(f"Received response from server {server_address[0]}:{server_address[1]}:")
+    print(f"-> Data: '{response}'")
+
+except socket.error as err:
+    print(f"Error during communication: {err}")
+
+finally:
+    # Close the socket
+    client_socket.close()
+    print("-" * 30)
+    print("Client finished and socket closed.")

udp_server.py

@@ -0,0 +1,58 @@
+import socket
+import sys
+
+# optionally take in some positional args for the port
+if len(sys.argv) > 1:
+    try:
+        PORT = int(sys.argv[1])
+    except ValueError:
+        print("Invalid port number. Using default port 12000.")
+        PORT = 12000
+else:
+    PORT = 12000
+
+# Define the server host and port
+HOST = '0.0.0.0'  # Standard loopback interface address (localhost)
+
+# Create a UDP socket
+try:
+    server_socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
+except socket.error as err:
+    print(f"Failed to create socket: {err}")
+    sys.exit()
+
+# Bind the socket to the address
+try:
+    server_socket.bind((HOST, PORT))
+    print(f"UDP Server listening on {HOST}:{PORT}")
+except socket.error as err:
+    print(f"Bind failed: {err}")
+    server_socket.close()
+    sys.exit()
+
+# Wait for and process incoming data
+while True:
+    try:
+        # Receive data and the client's address (buffer size 1024 bytes)
+        data, client_address = server_socket.recvfrom(1024)
+
+        # Decode the data and print the message
+        message = data.decode('utf-8')
+        print("-" * 30)
+        print(f"Received message from {client_address[0]}:{client_address[1]}:")
+        print(f"-> Data: '{message}'")
+
+        # Prepare the response message
+        response_message = f"Hello client! Server received: '{message.upper()}'"
+
+        # Send the response back to the client
+        server_socket.sendto(response_message.encode('utf-8'), client_address)
+        print(f"Sent response back to client.")
+
+    except Exception as e:
+        print(f"An error occurred: {e}")
+        break
+
+# Clean up (though usually unreachable in an infinite server loop)
+server_socket.close()
+print("Server stopped.")

updates/updates.go

@@ -119,7 +119,7 @@ func CheckForUpdate(owner, repo, currentVersion string) error {
 
 	// Check if update is available
 	if currentVer.isNewer(latestVer) {
-		printUpdateBanner(currentVer.String(), latestVer.String(), release.HTMLURL)
+		printUpdateBanner(currentVer.String(), latestVer.String(), "curl -fsSL https://static.pangolin.net/get-newt.sh | bash")
 	}
 
 	return nil
@@ -145,7 +145,7 @@ func printUpdateBanner(currentVersion, latestVersion, releaseURL string) {
 		"║  A newer version is available! Please update to get the" + padRight("", contentWidth-56) + "║",
 		"║  latest features, bug fixes, and security improvements." + padRight("", contentWidth-56) + "║",
 		emptyLine,
-		"║  Release URL: " + padRight(releaseURL, contentWidth-15) + "║",
+		"║  Update: " + padRight(releaseURL, contentWidth-10) + "║",
 		emptyLine,
 		borderBot,
 	}

util/util.go

@@ -33,6 +33,19 @@ func ResolveDomain(domain string) (string, error) {
 		port = ""
 	}
 
+	// Check if host is already an IP address (IPv4 or IPv6)
+	// For IPv6, the host from SplitHostPort will already have brackets stripped
+	// but if there was no port, we need to handle bracketed IPv6 addresses
+	cleanHost := strings.TrimPrefix(strings.TrimSuffix(host, "]"), "[")
+	if ip := net.ParseIP(cleanHost); ip != nil {
+		// It's already an IP address, no need to resolve
+		ipAddr := ip.String()
+		if port != "" {
+			return net.JoinHostPort(ipAddr, port), nil
+		}
+		return ipAddr, nil
+	}
+
 	// Lookup IP addresses
 	ips, err := net.LookupIP(host)
 	if err != nil {

websocket/client.go

@@ -46,6 +46,7 @@ type Client struct {
 	metricsCtxMu      sync.RWMutex
 	metricsCtx        context.Context
 	configNeedsSave   bool // Flag to track if config needs to be saved
+	serverVersion     string
 }
 
 type ClientOption func(*Client)
@@ -149,6 +150,10 @@ func (c *Client) GetConfig() *Config {
 	return c.config
 }
 
+func (c *Client) GetServerVersion() string {
+	return c.serverVersion
+}
+
 // Connect establishes the WebSocket connection
 func (c *Client) Connect() error {
 	go c.connectWithRetry()
@@ -351,9 +356,11 @@ func (c *Client) getToken() (string, error) {
 	}
 	defer resp.Body.Close()
 
+	body, _ := io.ReadAll(resp.Body)
+	logger.Debug("Token response body: %s", string(body))
+
 	if resp.StatusCode != http.StatusOK {
-		body, _ := io.ReadAll(resp.Body)
-		logger.Error("Failed to get token with status code: %d, body: %s", resp.StatusCode, string(body))
+		logger.Error("Failed to get token with status code: %d", resp.StatusCode)
 		telemetry.IncConnAttempt(ctx, "auth", "failure")
 		etype := "io_error"
 		if resp.StatusCode == http.StatusUnauthorized || resp.StatusCode == http.StatusForbidden {
@@ -368,7 +375,7 @@ func (c *Client) getToken() (string, error) {
 	}
 
 	var tokenResp TokenResponse
-	if err := json.NewDecoder(resp.Body).Decode(&tokenResp); err != nil {
+	if err := json.Unmarshal(body, &tokenResp); err != nil {
 		logger.Error("Failed to decode token response.")
 		return "", fmt.Errorf("failed to decode token response: %w", err)
 	}
@@ -381,6 +388,11 @@ func (c *Client) getToken() (string, error) {
 		return "", fmt.Errorf("received empty token from server")
 	}
 
+	// print server version
+	logger.Info("Server version: %s", tokenResp.Data.ServerVersion)
+
+	c.serverVersion = tokenResp.Data.ServerVersion
+
 	logger.Debug("Received token: %s", tokenResp.Data.Token)
 	telemetry.IncConnAttempt(ctx, "auth", "success")
 

websocket/types.go

@@ -9,7 +9,8 @@ type Config struct {
 
 type TokenResponse struct {
 	Data struct {
-		Token string `json:"token"`
+		Token         string `json:"token"`
+		ServerVersion string `json:"serverVersion"`
 	} `json:"data"`
 	Success bool   `json:"success"`
 	Message string `json:"message"`

wgtester/wgtester.go

@@ -38,7 +38,6 @@ type Server struct {
 	isRunning    bool
 	runningLock  sync.Mutex
 	newtID       string
-	outputPrefix string
 	useNetstack  bool
 	tnet         interface{} // Will be *netstack2.Net when using netstack
 }
@@ -50,7 +49,6 @@ func NewServer(serverAddr string, serverPort uint16, newtID string) *Server {
 		serverPort:   serverPort + 1, // use the next port for the server
 		shutdownCh:   make(chan struct{}),
 		newtID:       newtID,
-		outputPrefix: "[WGTester] ",
 		useNetstack:  false,
 		tnet:         nil,
 	}
@@ -63,7 +61,6 @@ func NewServerWithNetstack(serverAddr string, serverPort uint16, newtID string,
 		serverPort:   serverPort + 1, // use the next port for the server
 		shutdownCh:   make(chan struct{}),
 		newtID:       newtID,
-		outputPrefix: "[WGTester] ",
 		useNetstack:  true,
 		tnet:         tnet,
 	}
@@ -109,7 +106,7 @@ func (s *Server) Start() error {
 	s.isRunning = true
 	go s.handleConnections()
 
-	logger.Info("%sServer started on %s:%d", s.outputPrefix, s.serverAddr, s.serverPort)
+	logger.Debug("WGTester Server started on %s:%d", s.serverAddr, s.serverPort)
 	return nil
 }
 
@@ -127,7 +124,7 @@ func (s *Server) Stop() {
 		s.conn.Close()
 	}
 	s.isRunning = false
-	logger.Info("%sServer stopped", s.outputPrefix)
+	logger.Info("WGTester Server stopped")
 }
 
 // RestartWithNetstack stops the current server and restarts it with netstack
@@ -162,7 +159,7 @@ func (s *Server) handleConnections() {
 			// Set read deadline to avoid blocking forever
 			err := s.conn.SetReadDeadline(time.Now().Add(1 * time.Second))
 			if err != nil {
-				logger.Error("%sError setting read deadline: %v", s.outputPrefix, err)
+				logger.Error("Error setting read deadline: %v", err)
 				continue
 			}
 
@@ -192,7 +189,7 @@ func (s *Server) handleConnections() {
 					if err == io.EOF {
 						return
 					}
-					logger.Error("%sError reading from UDP: %v", s.outputPrefix, err)
+					logger.Error("Error reading from UDP: %v", err)
 				}
 				continue
 			}
@@ -224,7 +221,7 @@ func (s *Server) handleConnections() {
 			copy(responsePacket[5:13], buffer[5:13])
 
 			// Log response being sent for debugging
-			// logger.Debug("%sSending response to %s", s.outputPrefix, addr.String())
+			// logger.Debug("Sending response to %s", addr.String())
 
 			// Send the response packet - handle both regular UDP and netstack UDP
 			if s.useNetstack {
@@ -238,9 +235,9 @@ func (s *Server) handleConnections() {
 			}
 
 			if err != nil {
-				logger.Error("%sError sending response: %v", s.outputPrefix, err)
+				logger.Error("Error sending response: %v", err)
 			} else {
-				// logger.Debug("%sResponse sent successfully", s.outputPrefix)
+				// logger.Debug("Response sent successfully")
 			}
 		}
 	}