When something breaks,Ether already knows why.
Overview
Fleet health
38 of 38 agents online
All quiet.
Watching 38 hosts. Nothing needs you.
- Agents online
- 38 of 38
- Open incidents
- 0
- Events per minute
- 1,284
- edge-lb-0112ms
- prod-web-0136ms
- prod-web-0241ms
- prod-web-0338ms
- db-primary4ms
- worker-0227ms
Most tools wake you up to say something broke. Ether wakes you up with the reason and the fix.
Install it once.Read one answer.
01
Deploy
Run one command on each host
An agent is live in 30 seconds. It streams processes, network activity, file changes and system health to Ether.
02
Reason
Ether works out what is actually wrong
When an alert rule fires, Ether checks the host, correlates signals across your fleet and narrows a noisy night down to one cause.
03
Approve
You approve, then it acts
Ether proposes the fix and waits. You approve from the dashboard or from Slack, and every step lands in the audit log.
- Processes
- Network
- Files
- Health
$ curl -sSL https://releases.ethernullx.com/agent/install.sh | sudo bash -s -- --token <your-token>
Instead of every engineer running their own investigation with their own LLM, Ether does it once and the whole team sees the result.
Investigation 4812prod-web-03
ResolvedStarted by your alert rule
p95 latency above 2s
- 03:14:08Checked processesapi is holding 3,412 open connections
- 03:14:11Analyzed networkNo unusual traffic on this host
- 03:14:15Correlated fleetOnly hosts running api are affected
- 03:14:20Found the root causeA leaked connection pool in your API
- 03:15:41Restarted with a graceful drainApproved by Priya. p95 back to 37ms
Read this morning
- Priya Nair08:02
- Tomás Rivera08:06
- Wen Zhao08:11
Nobody had to run it again.
Used from your runbooks
- api restarts always drain first
- Disk spike at 02:00 is the nightly backup
It proposes.You decide.
Giving software a shell on production is a serious decision. So Ether asks first, every time, and writes down who said yes.
Restart a service
api on prod-web-03
Leaked connection pool
Block an IP
185.220.101.47 at edge-lb-01
412 failed SSH logins in 9 minutes
Isolate a host
worker-02
Unsigned binary calling out
Roll back a change
nginx.conf on edge-lb-01
Config edit preceded 502s by 40 seconds
Ether shows upwhere you already work.
Alerts, tickets and actions flow to the tools your team already uses.
- Slack
- Jira
- Telegram
- GitHub
- Google Workspace
- Any webhook
#incidents03:14
Ether
prod-web-03 is leaking connections in your API. A restart clears it.
Alerts and commands in your channels.
Before you give it a shell
Can Ether change anything without my approval?
No. Ether proposes a fix and waits. Nothing runs until someone on your team approves it, and the approval, the action and the result all go in the audit log.
What does the agent collect?
Processes, network activity, file changes and system health from the host it runs on. That stream is what Ether reads when it investigates.
How long does setup take?
You sign up, get an enrollment token and run one command on each host. The agent connects in about 30 seconds.
Is this for security or for reliability?
Both. A latency spike and a brute force attempt start the same way, as an alert nobody has context for. Ether investigates either one with the same evidence.
Does it replace my alerting tools?
You can define alert rules in Ethernull, then route findings to Slack, Jira, Telegram or any webhook. Your team keeps working where it works today.
Can I teach it how my systems work?
Yes. Add runbooks and describe expected behavior, like a disk spike at 02:00 being the nightly backup. Ether uses them in every investigation.
Which plan do I need?
Free covers one user and 10 agents. Pro covers 5 users and 50 agents, and adds Brain, playbooks and the world model. Enterprise has no limits and adds SSO, priority support and RedPhantom.
What is RedPhantom?
Authorized external penetration testing against targets you own. You add a target, describe what to test and read the result in the dashboard. It is part of the Enterprise plan.