Reference
Commands
Command surfaces referenced throughout the Mullgate documentation.
The documentation repeatedly anchors itself to the following CLI help contract:
mullgate --helpmullgate setup --helpmullgate proxy --helpmullgate proxy start --helpmullgate proxy logs --helpmullgate proxy relay --helpmullgate config --helpmullgate version --helpmullgate completions --help
Core workflow commands
| command | key flags | purpose |
|---|---|---|
mullgate setup | --non-interactive, --location, --exposure-mode | create or update canonical config and derived runtime artifacts |
mullgate proxy access | --mode, --access-mode, --base-domain, --route-bind-ip, --unsafe-public-empty-password | inspect or update exposure posture, access mode, host planning, DNS guidance, selector examples, and direct-IP entrypoints |
mullgate proxy export | --regions, --guided, selector flags | list region groups or generate client-ready proxy inventories in published-routes mode |
mullgate proxy relay list | selector flags such as --country, --owner, --provider | inspect relay candidates that match a policy |
mullgate proxy relay probe | --country, --count | latency-probe likely relay candidates |
mullgate proxy relay recommend | --country, --count, --apply | preview or pin an exact relay choice |
mullgate proxy relay verify | --route | verify a configured route across the published proxy protocols |
mullgate proxy validate | --refresh | refresh saved validation metadata and config-derived artifacts |
mullgate proxy start | --dry-run | render and validate artifacts, then optionally launch the runtime |
mullgate proxy stop | none | stop the saved Docker runtime bundle without rerendering artifacts |
mullgate proxy restart | none | stop the current runtime bundle, rerender artifacts, and start it again |
mullgate proxy status | none | inspect live runtime state |
mullgate proxy logs | --tail, --follow | read the saved Docker Compose logs for the current runtime bundle |
mullgate proxy doctor | none | diagnose routing, hostname, and runtime failures |
mullgate proxy autostart | enable, disable, status | manage Linux user-service startup, including linger-backed reboot startup |
mullgate config path | none | print config, state, cache, and runtime paths |
mullgate config show | none | print the saved canonical config |
mullgate config get | <key> | read one saved config value |
mullgate config set | <key> <value> | update one saved config value |
mullgate version | none | print CLI version plus support metadata |
mullgate completions <shell> | bash, zsh, fish | generate shell completion scripts |
Invocation forms
Mullgate is documented in two operator-facing forms:
- installed
mullgatecommand - packed release asset or extracted standalone binary
How the commands fit together
A practical workflow usually looks like this:
mullgate setupmullgate proxy access- if you want a proxy inventory file,
mullgate proxy export --regions - if you want a proxy inventory file,
mullgate proxy export --guided mullgate proxy relay listmullgate proxy relay probemullgate proxy relay recommendmullgate proxy accessmullgate proxy validate --refreshmullgate proxy start --dry-runmullgate proxy startmullgate proxy statusmullgate proxy logs --tail 50mullgate proxy doctormullgate proxy restartwhen you need a stop-plus-rerender cycle later
If you switch to inline-selector, skip mullgate proxy export and use mullgate proxy access to inspect the shared listener and selector examples. For the supported selector families and guaranteed URL shape, see Inline Selector Reference.
Reference practice
When in doubt, prefer the live CLI help output as the source of truth for flags, defaults, and command wording.