# Repository: trailofbits/algo # Stars: 30249 ## CLAUDE.md # CLAUDE.md - LLM Guidance for Algo VPN This document provides essential context and guidance for LLMs working on the Algo VPN codebase. ## Project Overview Algo is an Ansible-based tool that sets up a personal VPN in the cloud. It's designed to be: - **Security-focused**: Creates hardened VPN servers with minimal attack surface - **Easy to use**: Automated deployment with sensible defaults - **Multi-platform**: Supports various cloud providers and operating systems - **Privacy-preserving**: No logging, minimal data retention ### Core Technologies - **VPN Protocols**: WireGuard (preferred) and IPsec/IKEv2 - **Configuration Management**: Ansible (v12+) - **Languages**: Python, YAML, Shell, Jinja2 templates - **Supported Providers**: AWS, Azure, DigitalOcean, GCP, Vultr, Hetzner, local deployment ### Philosophy - Stability over features - Security over convenience - Clarity over cleverness - Test everything - Stay in scope - solve exactly what the issue asks, nothing more - Test assumptions - run the code before committing - Resist new dependencies - each one is attack surface and maintenance ## Architecture and Structure ``` algo/ ├── main.yml # Primary playbook ├── users.yml # User management playbook ├── server.yml # Server-specific tasks ├── config.cfg # Main configuration file ├── pyproject.toml # Python project configuration and dependencies ├── uv.lock # Exact dependency versions lockfile ├── requirements.yml # Ansible collections ├── roles/ # Ansible roles │ ├── common/ # Base system configuration, firewall, hardening │ ├── wireguard/ # WireGuard VPN setup │ ├── strongswan/ # IPsec/IKEv2 setup │ ├── dns/ # DNS configuration (dnscrypt-proxy) │ └── cloud-*/ # Cloud provider specific roles ├── library/ # Custom Ansible modules └── tests/unit/ # Python unit tests ``` ## Development Workflow ### Quality Gates (MANDATORY) **All PRs must pass these checks locally before submission.** CI will reject failures: ```bash # Run the full lint suite (same as CI) ansible-lint . && yamllint . && ruff check . && shellcheck scripts/*.sh && semgrep --config auto --exclude-rule dockerfile.security.last-user-is-root.last-user-is-root --error --quiet . ansible-playbook main.yml --syntax-check ansible-playbook users.yml --syntax-check pytest tests/unit/ -q ``` Common lint issues to fix before submitting: - YAML files missing `---` document start markers - GitHub workflows with unquoted `on:` (must be `'on':`) - Using `ignore_errors: true` instead of `failed_when: false` - Jinja2 spacing errors (`{{foo}}` should be `{{ foo }}`) - Missing `mode:` on file/directory tasks ### Zero-Tolerance Warning Policy **No warnings are tolerated in CI.** Every linter finding must be either fixed or explicitly allowlisted in the tool's config file (`.ansible-lint`, `pyproject.toml`, etc.). Why this matters for Algo: - **Security tool** - VPN misconfigurations silently break privacy guarantees. A "cosmetic" warning today hides a real bug tomorrow. - **Ansible complexity** - YAML+Jinja2 linting catches real runtime failures (wrong key order breaks `when` evaluation, spacing errors cause template failures). Warnings in Ansible are not style nits. - **CI signal integrity** - If 30 warnings scroll by on every run, the 31st one (a real regression) goes unnoticed. Zero warnings means every new finding gets human attention. Resolution order of preference: 1. **Fix it** - Preferred. Most findings have straightforward fixes. 2. **Allowlist in config** - If the rule is wrong for this project, add to `skip_list` with a comment explaining why. 3. **Inline suppress** - Last resort. Use `# noqa: rule-name` with a comment justifying the exception. Never use `warn_list` in `.ansible-lint` — it exists as a migration tool, not a permanent home. Rules either pass or are explicitly skipped. ### Design Requirements When adding or modifying features, verify these before requesting review: 1. **Validate inputs early** - Check for empty lists, missing configs, permission mismatches before expensive operations 2. **Explicit file modes** - Always specify `mode:` on file/directory tasks (never rely on umask) 3. **Fail vs warn** - Permission/security issues should fail; optional features can warn 4. **Actionable errors** - Include fix commands in error messages: `"Run: sudo chown -R $USER configs/"` 5. **Follow existing patterns** - Search codebase first: `rg "when:.*localhost" --type yaml` ### Linting Tools | Tool | Target | Key Rules | |------|--------|-----------| | `ansible-lint` | YAML tasks | Use `failed_when` not `ignore_errors`, add `mode:` to files | | `yamllint` | All YAML | Document start `---`, quote `'on':` in workflows | | `ruff` | Python | Line length 120, target Python 3.11 | | `shellcheck` | Shell scripts | Quote variables, use `set -euo pipefail` | | `semgrep` | All code | SAST scanner, `--config auto`, suppress with `# nosemgrep: rule-id` | ### Git Workflow 1. Create feature branches from `master` 2. Run all linters before pushing 3. Make atomic commits with clear messages 4. Update PR description with test results ### Self-Review Checklist Before creating a PR, review your own diff: - [ ] Did I run all linters locally? - [ ] Did I search for similar patterns in the codebase? - [ ] Did I add explicit `mode:` to file/directory tasks? - [ ] Did I validate inputs before expensive operations? - [ ] Did I update tests if I changed file paths or behavior? - [ ] Would a reviewer ask "what happens if X is empty/missing?" ## Ansible Pitfalls ### with_items vs loop `with_items` auto-flattens lists; `loop` does not. **Never mechanically convert:** ```yaml # WRONG - treats list as single item, creates file named "['alice', 'bob']" loop: - "{{ users }}" # CORRECT - iterates over list contents loop: "{{ users }}" # CORRECT - combining lists (with_items did this automatically) loop: "{{ users + [server_name] }}" ``` **Always test loop conversions** - verify the task creates expected files. ### Path Variables Never include trailing slashes - causes double-slash bugs: ```yaml # WRONG - creates paths like /etc/ipsec.d//private ipsec_path: "configs/{{ server }}/ipsec/" # CORRECT ipsec_path: "configs/{{ server }}/ipsec" ``` ### ignore_errors vs failed_when ```yaml # WRONG - ansible-lint failure - name: Clear history command: some_command ignore_errors: true # CORRECT - explicit about expected failures - name: Clear history command: some_command failed_when: false ``` ### changed_when on Read-Only Tasks Handlers and check commands that don't modify state need `changed_when: false`: ```yaml - name: Check service status command: systemctl status foo changed_when: false ``` ### Jinja2 Native Mode (Ansible 12+) Ansible 12 enables `jinja2_native` by default, changing how values are evaluated: **Boolean conditionals require actual booleans:** ```yaml # WRONG - string "true" is not boolean ipv6_support: "{% if ipv6 %}true{% else %}false{% endif %}" # CORRECT - return actual boolean ipv6_support: "{{ ipv6 is defined }}" ``` **No nested templates in lookup():** ```yaml # WRONG - deprecated double-templating key: "{{ lookup('file', '{{ SSH_keys.public }}') }}" # CORRECT - pass variable directly key: "{{ lookup('file', SSH_keys.public) }}" ``` **JSON files need explicit parsing:** ```yaml # WRONG - returns string in native mode creds: "{{ lookup('file', 'credentials.json') }}" # CORRECT - parse JSON explicitly creds: "{{ lookup('file', 'credentials.json') | from_json }}" ``` **default() doesn't trigger on empty strings:** ```yaml # WRONG - empty string '' is not undefined key: "{{ lookup('env', 'AWS_KEY') | default('fallback') }}" # CORRECT - add true to handle falsy values key: "{{ lookup('env', 'AWS_KEY') | default('fallback', true) }}" ``` **Complex Jinja loops break in set_fact:** ```yaml # WRONG - list comprehension fails in native mode servers: "[{% for s in configs %}{{ s.name }},{% endfor %}]" # CORRECT - use Ansible loop servers: "{{ servers | default([]) + [item.name] }}" loop: "{{ configs }}" ``` **Use tests (not filters) for boolean checks:** ```yaml # WRONG - filters return transformed data, not booleans that: my_ip | ansible.utils.ipv4 # CORRECT - tests return native booleans that: my_ip is ansible.utils.ipv4_address ``` ## DNS Architecture Algo uses a randomly generated IP in 172.16.0.0/12 on the loopback interface (`local_service_ip`) for DNS. This provides consistency across WireGuard and IPsec but requires understanding systemd socket activation. ### Why This Design - Consistent DNS IP across both VPN protocols - Survives interface changes and restarts - Works identically across all cloud providers - Trade-off: Requires `route_localnet=1` sysctl ### systemd Socket Activation Ubuntu's dnscrypt-proxy uses socket activation which **completely ignores** the `listen_addresses` config setting. You must configure the socket, not the service: ```ini # /etc/systemd/system/dnscrypt-proxy.socket.d/10-algo-override.conf [Socket] ListenStream= # Clear defaults first ListenDatagram= ListenStream=172.x.x.x:53 # Then set VPN IP ListenDatagram=172.x.x.x:53 ``` Common mistakes: - Trying to disable/mask the socket (breaks service dependency) - Only setting ListenStream (need ListenDatagram for UDP) - Forgetting to restart socket after config changes ### Debugging DNS Many "routing" issues are actually DNS issues. Start here: ```bash ss -lnup | grep :53 # Should show local_service_ip:53 systemctl status dnscrypt-proxy.socket # Check for config warnings sysctl net.ipv4.conf.all.route_localnet # Must be 1 dig @172.x.x.x google.com # Test resolution ``` For comprehensive diagnostics, see [docs/troubleshooting.md](docs/troubleshooting.md#diagnostic-commands). ## Common Issues ### iptables Backend (nft vs legacy) Ubuntu 22.04+ defaults to iptables-nft which reorders rules unpredictably. Algo forces iptables-legacy for consistent behavior. Switching backends can break DNS routing that previously worked. ### Multi-homed Systems (DigitalOcean, etc.) Servers with both public and private IPs on the same interface need explicit output interface for NAT: ```yaml -o {{ ansible_default_ipv4['interface'] }} ``` Don't overengineer with SNAT - MASQUERADE with interface specification works fine. ### OpenSSL Version Compatibility OpenSSL 3.x dropped support for legacy algorithms. Add `-legacy` flag conditionally: ```yaml {{ (openssl_version is version('3', '>=')) | ternary('-legacy', '') }} ``` ### IPv6 Endpoint Formatting WireGuard configs must bracket IPv6 addresses: ```jinja2 {% if ':' in IP %}[{{ IP }}]:{{ port }}{% else %}{{ IP }}:{{ port }}{% endif %} ``` ### Jinja2 Templates Many templates use Ansible-specific filters. Test with `tests/unit/test_template_rendering.py` and mock Ansible filters when testing. ## Time Wasters to Avoid Lessons learned - don't spend time on these unless absolutely necessary: 1. **Converting MASQUERADE to SNAT** - MASQUERADE works fine for Algo's use case 2. **Fighting systemd socket activation** - Configure it properly instead of disabling 3. **Debugging NAT before checking DNS** - Most "routing" issues are DNS issues 4. **Complex IPsec policy matching** - Keep NAT rules simple 5. **Testing on existing servers** - Always test on fresh deployments 6. **Interface-specific route_localnet** - WireGuard interface doesn't exist until service starts 7. **DNAT for loopback addresses** - Packets to local IPs don't traverse PREROUTING ## What to Avoid - **Speculative features** - Don't add "might be useful" functionality. Open an issue instead. - **New dependencies without justification** - Vanilla Ansible/Python can do most things. - **Bundling unrelated fixes** - One PR, one purpose. Separate issues get separate PRs. - **Assuming behavior** - If converting `with_items` to `loop`, test that it still works. If adding a firewall rule, verify packets flow. - **Configuration options** - Don't add flags unless users actively need them. Each option doubles testing surface. - **Undocumented workarounds** - When working around broken upstream modules, file an issue and add a comment linking to it. Future maintainers need to know why workarounds exist. ## Writing Effective Tests When writing tests, **verify your test actually detects the failure case** (mutation testing approach): 1. Write the test for the bug you're preventing 2. Temporarily introduce the bug to verify the test fails 3. Fix the bug and verify the test passes 4. Document what specific issue the test prevents ```python def test_regression_openssl_inline_comments(): """Tests that we detect inline comments in Jinja2 expressions.""" # This pattern SHOULD fail (has inline comments) problematic = "{{ ['DNS:' + id, # comment ] }}" assert not validate(problematic), "Should detect inline comments" # This pattern SHOULD pass (no inline comments) fixed = "{{ ['DNS:' + id] }}" assert validate(fixed), "Should pass without comments" ``` ## Quick Reference ### Local Development Setup ```bash uv sync uv run ansible-galaxy install -r requirements.yml ansible-playbook main.yml -e "provider=local" ``` ### Common Commands ```bash # Add/update users ansible-playbook users.yml -e "server=SERVER_NAME" # Update dependencies uv lock && pytest tests/unit/ -q # Debug deployment ansible-playbook main.yml -vvv ``` ### Key Directories - `configs/` - Generated client configurations - `roles/*/tasks/` - Main task files - `roles/*/templates/` - Jinja2 templates - `library/` - Custom Ansible modules (add to `mock_modules` in `.ansible-lint`) ## Non-Interactive Deployment All `pause:` prompts in `input.yml` and provider roles skip when their variable is pre-defined via `-e` or environment variables. This enables fully headless deployment for CI, agents, and scripted workflows. See [docs/deploy-from-ansible.md](docs/deploy-from-ansible.md) for full human-facing documentation. ### Core variables These bypass the main prompts in `input.yml`: | Variable | Type | Default | Purpose | |----------|------|---------|---------| | `provider` | string | *(prompt)* | Provider alias (e.g., `digitalocean`, `ec2`, `local`) | | `server_name` | string | `algo` | VPN server name | | `ondemand_cellular` | bool | `false` | iOS/macOS Connect On Demand for cellular | | `ondemand_wifi` | bool | `false` | iOS/macOS Connect On Demand for Wi-Fi | | `ondemand_wifi_exclude` | string | *(none)* | Comma-separated trusted Wi-Fi networks | | `store_pki` | bool | `false` | Retain PKI keys (needed to add users later) | | `dns_adblocking` | bool | `false` | Enable DNS ad blocking | | `ssh_tunneling` | bool | `false` | Per-user SSH tunnel accounts | ### Provider credentials | Provider | `-e` variables | Env var fallbacks | |----------|---------------|-------------------| | `digitalocean` | `do_token`, `region` | `DO_API_TOKEN` | | `ec2` | `aws_access_key`, `aws_secret_key`, `region` | `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY` (also reads `~/.aws/credentials`) | | `lightsail` | `aws_access_key`, `aws_secret_key`, `region` | `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY` | | `azure` | `azure_secret`, `azure_tenant`, `azure_client_id`, `azure_subscription_id`, `region` | `AZURE_SECRET`, `AZURE_TENANT`, `AZURE_CLIENT_ID`, `AZURE_SUBSCRIPTION_ID` | | `gce` | `gce_credentials_file`, `region` | `GCE_CREDENTIALS_FILE_PATH` | | `hetzner` | `hcloud_token`, `region` | `HCLOUD_TOKEN` | | `vultr` | `vultr_config`, `region` | `VULTR_API_CONFIG` | | `scaleway` | `scaleway_token`, `scaleway_org_id`, `region` | `SCW_TOKEN`, `SCW_DEFAULT_ORGANIZATION_ID` | | `linode` | `linode_token`, `region` | `LINODE_API_TOKEN` | | `cloudstack` | `cs_key`, `cs_secret`, `cs_url`, `region` | `CLOUDSTACK_KEY`, `CLOUDSTACK_SECRET`, `CLOUDSTACK_ENDPOINT` | | `openstack` | `region` | `OS_AUTH_URL` (source your `openrc.sh`) | | `local` | `server`, `endpoint`, `local_install_confirmed` | *(none)* | ### Minimal examples ```bash # DigitalOcean — fully headless ansible-playbook main.yml -e \ "provider=digitalocean server_name=algo region=nyc3 do_token=YOUR_TOKEN ondemand_cellular=false ondemand_wifi=false dns_adblocking=false ssh_tunneling=false store_pki=false" # Local — for CI/testing ansible-playbook main.yml -e \ "provider=local server=localhost endpoint=10.0.0.1 local_install_confirmed=true ondemand_cellular=false ondemand_wifi=false dns_adblocking=false ssh_tunneling=false" ``` ### Updating users non-interactively ```bash ansible-playbook users.yml -e "server=YOUR_SERVER ca_password=YOUR_CA_PASS" ``` The `server` variable bypasses the server selection prompt. `ca_password` is only required when IPsec is enabled. ## Security Considerations - **Never expose secrets** - No passwords/keys in commits - **CVE Response** - Update immediately when security issues found - **Least Privilege** - Minimal permissions, dropped capabilities - **Secure Defaults** - Strong crypto (secp384r1), no logging, strict firewall ## Platform Support - **Primary OS**: Ubuntu 22.04/24.04 LTS - **Secondary**: Debian 11/12 - **Architectures**: x86_64 and ARM64 - **Testing tip**: DigitalOcean droplets have both public and private IPs on eth0, making them good test cases for multi-IP NAT scenarios ## README.md # Algo VPN [![Twitter](https://img.shields.io/twitter/url/https/twitter.com/fold_left.svg?style=social&label=Follow%20%40AlgoVPN)](https://x.com/AlgoVPN) Algo VPN is a set of Ansible scripts that simplify the setup of a personal WireGuard and IPsec VPN. It uses the most secure defaults available and works with common cloud providers. See our [release announcement](https://blog.trailofbits.com/2016/12/12/meet-algo-the-vpn-that-works/) for more information. ## Features * Supports only IKEv2 with strong crypto (AES-GCM, SHA2, and P-256) for iOS, MacOS, and Linux * Supports [WireGuard](https://www.wireguard.com/) for all of the above, in addition to Android and Windows 11 * Generates .conf files and QR codes for iOS, macOS, Android, and Windows WireGuard clients * Generates Apple profiles to auto-configure iOS and macOS devices for IPsec - no client software required * Includes helper scripts to add, remove, and manage users * Blocks ads with a local DNS resolver (optional) * Sets up limited SSH users for tunneling traffic (optional) * Privacy-focused with minimal logging, automatic log rotation, and configurable privacy enhancements * Based on Ubuntu 22.04 LTS with automatic security updates * Installs to DigitalOcean, Amazon Lightsail, Amazon EC2, Vultr, Microsoft Azure, Google Compute Engine, Scaleway, OpenStack, CloudStack, Hetzner Cloud, Linode, or [your own Ubuntu server (for advanced users)](docs/deploy-to-ubuntu.md) ## Anti-features * Does not support legacy cipher suites or protocols like L2TP, IKEv1, or RSA * Does not install Tor, OpenVPN, or other risky servers * Does not depend on the security of [TLS](https://tools.ietf.org/html/rfc7457) * Does not claim to provide anonymity or censorship avoidance * Does not claim to protect you from the [FSB](https://en.wikipedia.org/wiki/Federal_Security_Service), [MSS](https://en.wikipedia.org/wiki/Ministry_of_State_Security_(China)), [DGSE](https://en.wikipedia.org/wiki/Directorate-General_for_External_Security), or [FSM](https://en.wikipedia.org/wiki/Flying_Spaghetti_Monster) ## Deploy the Algo Server The easiest way to get an Algo server running is to run it on your local system or from [Google Cloud Shell](docs/deploy-from-cloudshell.md) and let it set up a _new_ virtual machine in the cloud for you. 1. **Setup an account on a cloud hosting provider.** Algo supports [DigitalOcean](https://m.do.co/c/4d7f4ff9cfe4) (most user friendly), [Amazon Lightsail](https://aws.amazon.com/lightsail/), [Amazon EC2](https://aws.amazon.com/), [Vultr](https://www.vultr.com/), [Microsoft Azure](https://azure.microsoft.com/), [Google Compute Engine](https://cloud.google.com/compute/), [Scaleway](https://www.scaleway.com/), [DreamCompute](https://www.dreamhost.com/cloud/computing/), [Linode](https://www.linode.com), other OpenStack-based cloud hosting, CloudStack-based cloud hosting, or [Hetzner Cloud](https://www.hetzner.com/). 2. **Get a copy of Algo.** The Algo scripts will be run from your local system. There are two ways to get a copy: - Download the [ZIP file](https://github.com/trailofbits/algo/archive/master.zip). Unzip the file to create a directory named `algo-master` containing the Algo scripts. - Use `git clone` to create a directory named `algo` containing the Algo scripts: ```bash git clone https://github.com/trailofbits/algo.git ``` 3. **Set your configuration options.** Open `config.cfg` in your favorite text editor. Specify the users you want to create in the `users` list. Create a unique user for each device you plan to connect to your VPN. You should also review the other options before deployment, as changing your mind about them later [may require you to deploy a brand new server](https://github.com/trailofbits/algo/blob/master/docs/faq.md#i-deployed-an-algo-server-can-you-update-it-with-new-features). 4. **Start the deployment.** Return to your terminal. In the Algo directory, run the appropriate script for your platform: **macOS/Linux:** ```bash ./algo ``` **Windows:** ```powershell .\algo.ps1 ``` The first time you run the script, it will automatically install the required Python environment (Python 3.11+). On subsequent runs, it starts immediately and works on all platforms (macOS, Linux, Windows via WSL). The Windows PowerShell script automatically uses WSL when needed, since Ansible requires a Unix-like environment. There are several optional features available, none of which are required for a fully functional VPN server. These optional features are described in the [deployment documentation](docs/deploy-from-ansible.md). That's it! You can now set up clients to connect to your VPN. Proceed to [Configure the VPN Clients](#configure-the-vpn-clients) below. ``` "# Congratulations! #" "# Your Algo server is running. #" "# Config files and certificates are in the ./configs/ directory. #" "# Go to https://whoer.net/ after connecting #" "# and ensure that all your traffic passes through the VPN. #" "# Local DNS resolver 172.16.0.1 #" "# The p12 and SSH keys password for new users is XXXXXXXX #" "# The CA key password is XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX #" "# Shell access: ssh -F configs//ssh_config #" ``` ## Configure the VPN Clients Certificates and configuration files that users will need are placed in the `configs` directory. Make sure to secure these files since many contain private keys. All files are saved under a subdirectory named with the IP address of your new Algo VPN server. **Important for IPsec users**: If you want to add or delete users later, you must select `yes` at the `Do you want to retain the keys (PKI)?` prompt during the server deployment. This preserves the certificate authority needed for user management. ### Apple WireGuard is used to provide VPN services on Apple devices. Algo generates a WireGuard configuration file, `wireguard/.conf`, and a QR code, `wireguard/.png`, for each user defined in `config.cfg`. On iOS, install the [WireGuard](https://itunes.apple.com/us/app/wireguard/id1441195209?mt=8) app from the iOS App Store. Then, use the WireGuard app to scan the QR code or AirDrop the configuration file to the device. On macOS, install the [WireGuard](https://itunes.apple.com/us/app/wireguard/id1451685025?mt=12) app from the Mac App Store. WireGuard will appear in the menu bar once you run the app. Click on the WireGuard icon, choose **Import tunnel(s) from file...**, then select the appropriate WireGuard configuration file. On either iOS or macOS, you can enable "Connect on Demand" and/or exclude certain trusted Wi-Fi networks (such as your home or work) by editing the tunnel configuration in the WireGuard app. (Algo can't do this automatically for you.) If you prefer to use the built-in IPsec VPN on Apple devices, or need "Connect on Demand" or excluded Wi-Fi networks automatically configured, see the [Apple IPsec client setup guide](docs/client-apple-ipsec.md) for detailed configuration instructions. ### Android WireGuard is used to provide VPN services on Android. Install the [WireGuard VPN Client](https://play.google.com/store/apps/details?id=com.wireguard.android). Import the corresponding `wireguard/.conf` file to your device, then set up a new connection with it. See the [Android setup guide](docs/client-android.md) for detailed installation and configuration instructions. ### Windows WireGuard is used to provide VPN services on Windows. Algo generates a WireGuard configuration file, `wireguard/.conf`, for each user defined in `config.cfg`. Install the [WireGuard VPN Client](https://www.wireguard.com/install/#windows-7-8-81-10-2012-2016-2019). Import the generated `wireguard/.conf` file to your device, then set up a new connection with it. See the [Windows setup instructions](docs/client-windows.md) for more detailed walkthrough and troubleshooting. ### Linux Linux clients can use either WireGuard or IPsec: WireGuard: WireGuard works great with Linux clients. See the [Linux WireGuard setup guide](docs/client-linux-wireguard.md) for step-by-step instructions on configuring WireGuard on Ubuntu and other distributions. IPsec: For strongSwan IPsec clients (including OpenWrt, Ubuntu Server, and other distributions), see the [Linux IPsec setup guide](docs/client-linux-ipsec.md) for detailed configuration instructions. ### OpenWrt For OpenWrt routers using WireGuard, see the [OpenWrt WireGuard setup guide](docs/client-openwrt-router-wireguard.md) for router-specific configuration instructions. ### Other Devices For devices not covered above or manual configuration, you'll need specific certificate and configuration files. The files you need depend on your device platform and VPN protocol (WireGuard or IPsec). * ipsec/manual/cacert.pem: CA Certificate * ipsec/manual/.p12: User Certificate and Private Key (in PKCS#12 format) * ipsec/manual/.conf: strongSwan client configuration * ipsec/manual/.secrets: strongSwan client configuration * ipsec/apple/.mobileconfig: Apple Profile * wireguard/.conf: WireGuard configuration profile * wireguard/.png: WireGuard configuration QR code ## Setup an SSH Tunnel If you turned on the optional SSH tunneling role, local user accounts will be created for each user in `config.cfg`, and SSH authorized_key files for them will be in the `configs` directory (user.pem). SSH user accounts do not have shell access, cannot authenticate with a password, and only have limited tunneling options (e.g., `ssh -N` is required). This ensures that SSH users have the least access required to set up a tunnel and can perform no other actions on the Algo server. Use the example command below to start an SSH tunnel by replacing `` and `` with your own. Once the tunnel is set up, you can configure a browser or other application to use 127.0.0.1:1080 as a SOCKS proxy to route traffic through the Algo server: ```bash ssh -D 127.0.0.1:1080 -f -q -C -N @algo -i configs//ssh-tunnel/.pem -F configs//ssh_config ``` ## SSH into Algo Server Your Algo server is configured for key-only SSH access for administrative purposes. Open the Terminal app, `cd` into the `algo-master` directory where you originally downloaded Algo, and then use the command listed on the success message: ``` ssh -F configs//ssh_config ``` where `` is the IP address of your Algo server. If you find yourself regularly logging into the server, it will be useful to load your Algo SSH key automatically. Add the following snippet to the bottom of `~/.bash_profile` to add it to your shell environment permanently: ``` ssh-add ~/.ssh/algo > /dev/null 2>&1 ``` Alternatively, you can choose to include the generated configuration for any Algo servers created into your SSH config. Edit the file `~/.ssh/config` to include this directive at the top: ``` Include /configs/*/ssh_config ``` where `` is the directory where you cloned Algo. ## Adding or Removing Users Algo makes it easy to add or remove users from your VPN server after initial deployment. For IPsec users: You must have selected `yes` at the `Do you want to retain the keys (PKI)?` prompt during the initial server deployment. This preserves the certificate authority needed for user management. You should also save the p12 and CA key passwords shown during deployment, as they're only displayed once. To add or remove users, first edit the `users` list in your `config.cfg` file. Add new usernames or remove existing ones as needed. Then navigate to the algo directory in your terminal and run: **macOS/Linux:** ```bash ./algo update-users ``` **Windows:** ```powershell .\algo.ps1 update-users ``` After the process completes, new configuration files will be generated in the `configs` directory for any new users. The Algo VPN server will be updated to contain only the users listed in the `config.cfg` file. Removed users will no longer be able to connect, and new users will have fresh certificates and configuration files ready for use. ## Privacy and Logging Algo takes a pragmatic approach to privacy. By default, we minimize logging while maintaining enough information for security and troubleshooting. What IS logged by default: * System security events (failed SSH attempts, firewall blocks, system updates) * Kernel messages and boot diagnostics (with reduced verbosity) * WireGuard client state (visible via `sudo wg` - shows last endpoint and handshake time) * Basic service status (service starts/stops/errors) * All logs automatically rotate and delete after 7 days Privacy is controlled by two main settings in `config.cfg`: * `strongswan_log_level: -1` - Controls StrongSwan connection logging (-1 = disabled, 2 = debug) * `privacy_enhancements_enabled: true` - Master switch for log rotation, history clearing, log filtering, and cleanup To enable full debugging when troubleshooting, set both `strongswan_log_level: 2` and `privacy_enhancements_enabled: false`. This will capture detailed connection logs and disable all privacy features. Remember to revert these changes after debugging. After deployment, verify your privacy settings: ```bash ssh -F configs//ssh_config sudo /usr/local/bin/privacy-monitor.sh ``` Perfect privacy is impossible with any VPN solution. Your cloud provider sees and logs network traffic metadata regardless of your server configuration. And of course, your ISP knows you're connecting to a VPN server, even if they can't see what you're doing through it. For the highest level of privacy, treat your Algo servers as disposable. Spin up a new instance when you need it, use it for your specific purpose, then destroy it completely. The ephemeral nature of cloud infrastructure can be a privacy feature if you use it intentionally. ## Additional Documentation * [FAQ](docs/faq.md) * [Troubleshooting](docs/troubleshooting.md) * How Algo uses [Firewalls](docs/firewalls.md) ### Setup Instructions for Specific Cloud Providers * Configure [Amazon EC2](docs/cloud-amazon-ec2.md) * Configure [Azure](docs/cloud-azure.md) * Configure [DigitalOcean](docs/cloud-do.md) * Configure [Google Cloud Platform](docs/cloud-gce.md) * Configure [Vultr](docs/cloud-vultr.md) * Configure [CloudStack](docs/cloud-cloudstack.md) * Configure [Hetzner Cloud](docs/cloud-hetzner.md) ### Install and Deploy from Common Platforms * Deploy from [macOS](docs/deploy-from-macos.md) * Deploy from [Windows](docs/deploy-from-windows.md) * Deploy from [Google Cloud Shell](docs/deploy-from-cloudshell.md) * Deploy from a [Docker container](docs/deploy-from-docker.md) ### Setup VPN Clients to Connect to the Server * Setup [Windows](docs/client-windows.md) clients * Setup [Android](docs/client-android.md) clients * Setup [Linux](docs/client-linux.md) clients with Ansible * Setup Ubuntu clients to use [WireGuard](docs/client-linux-wireguard.md) * Setup Linux clients to use [IPsec](docs/client-linux-ipsec.md) * Setup Apple devices to use [IPsec](docs/client-apple-ipsec.md) * Setup Macs running macOS 10.13 or older to use [WireGuard](docs/client-macos-wireguard.md) ### Advanced Deployment * Deploy to your own [Ubuntu](docs/deploy-to-ubuntu.md) server, and road warrior setup * Deploy from [Ansible](docs/deploy-from-ansible.md) non-interactively * Deploy onto a [cloud server at time of creation with shell script or cloud-init](docs/deploy-from-script-or-cloud-init-to-localhost.md) * Deploy to an [unsupported cloud provider](docs/deploy-to-unsupported-cloud.md) If you've read all the documentation and have further questions, [create a new discussion](https://github.com/trailofbits/algo/discussions). ## Endorsements > I've been ranting about the sorry state of VPN svcs for so long, probably about > time to give a proper talk on the subject. TL;DR: use Algo. -- [Kenn White](https://twitter.com/kennwhite/status/814166603587788800) > Before picking a VPN provider/app, make sure you do some research > https://research.csiro.au/ng/wp-content/uploads/sites/106/2016/08/paper-1.pdf ... – or consider Algo -- [The Register](https://twitter.com/TheRegister/status/825076303657177088) > Algo is really easy and secure. -- [the grugq](https://twitter.com/thegrugq/status/786249040228786176) > I played around with Algo VPN, a set of scripts that let you set up a VPN in the cloud in very little time, even if you don’t know much about development. I’ve got to say that I was quite impressed with Trail of Bits’ approach. -- [Romain Dillet](https://twitter.com/romaindillet/status/851037243728965632) for [TechCrunch](https://techcrunch.com/2017/04/09/how-i-made-my-own-vpn-server-in-15-minutes/) > If you’re uncomfortable shelling out the cash to an anonymous, random VPN provider, this is the best solution. -- [Thorin Klosowski](https://twitter.com/kingthor) for [Lifehacker](http://lifehacker.com/how-to-set-up-your-own-completely-free-vpn-in-the-cloud-1794302432) ## Contributing See our [Development Guide](docs/DEVELOPMENT.md) for information on: * Setting up your development environment * Using prek hooks for code quality * Running tests and linters * Contributing code via pull requests ## Support Algo VPN [![PayPal](https://www.paypalobjects.com/en_US/i/btn/btn_donate_SM.gif)](https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=CYZZD39GXUJ3E) [![Patreon](https://img.shields.io/badge/back_on-patreon-red.svg)](https://www.patreon.com/algovpn) All donations support continued development. Thanks! * We accept donations via [PayPal](https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=CYZZD39GXUJ3E) and [Patreon](https://www.patreon.com/algovpn). * Use our [referral code](https://m.do.co/c/4d7f4ff9cfe4) when you sign up to Digital Ocean for a $10 credit. * We also accept and appreciate contributions of new code and bugfixes via Github Pull Requests. Algo is licensed and distributed under the AGPLv3. If you want to distribute a closed-source modification or service based on Algo, then please consider purchasing an exception . As with the methods above, this will help support continued development.