Think inside your AI world.

Prometheus through Unl

Prometheus's MCP server will tell you exactly which target is down and for how long. Whether that's a real outage or a broken scrape path is a distinction your own escalation rule draws - a red target isn't automatically a page.

Prometheus through Unl reads live target health and firing alerts against the real-outage rule you have ratified, so 'is this actually an outage or a broken scrape?' comes back unconfirmed with the condition it hasn't met.

What Prometheus holds

Query Prometheus's MCP server about a target or alert and it returns:

  • instant and range PromQL query results
  • target discovery state, showing which targets are up or down
  • all currently firing alerts and loaded alerting or recording rules
  • label names and values for a given time range and matchers
  • metadata about metrics currently being scraped

What the naked read gives you

Prometheus's MCP server runs the instant or range PromQL query, lists which targets are up or down across service discovery, lists every firing alert and loaded rule, and returns metric metadata and label values on request. It reports exactly what the monitoring layer is seeing right now, target by target, without saying whether what it's seeing is a real outage or a scraping problem.

The frame judges the data it is given; it does not verify the source’s accuracy.

What changes when Prometheus is read measured

An SRE at a logistics platform ratified an escalation rule in Unl: a 'service down' alert only pages on-call once the target itself reports a genuine scrape failure for more than five minutes, because a firewall change during a network migration once made a target briefly unreachable and paged on-call at 2am for an outage that never actually happened to the service.

'Is this actually an outage or a broken scrape?' returns, measured: unconfirmed - the target read shows the checkout-service target unreachable for six minutes, but a query against up{job="checkout-service"} shows healthy responses via a second scrape path, pointing to a routing issue rather than a real outage. Prometheus supplied the target and query state; your ratified rule supplied the confirmation bar.

And back again

Whichever way this resolves - the second path drops too, or the first one recovers - Prometheus through Unl catches the settled target state on its next look. The rule then either confirms the page or stands it down, right here.

The answer comes back measured against what you already decided, and why.

The lane is live and open to this tool today: one box, paste anything. If it speaks MCP, Unl can reach it. Readings arrive unprompted, the data beside the criterion; Unl is a courier, not a warehouse, and keeps only your keys and the frame.

Questions people ask

What would confirm this as a real outage?

Both scrape paths to the checkout service reporting the target down at once, rather than just one. The same ratified rule reads that combined state and confirms the page.

How do I connect Prometheus to Claude?

Run the official Prometheus MCP server (prometheus/prometheus-mcp) against your Prometheus instance's API endpoint, then register it in Claude's MCP configuration - Unl's reads travel through that same connection, with destructive TSDB admin tools left disabled by default.

Does Unl write to my Prometheus instance?

Unl reads through Prometheus, and can write back on your explicit gesture — it never acts as a side effect of a read.

What this is

Think inside your AI world — you stay in command

Unlimitless (Unl to friends) holds what you've settled, reads what your tools are showing, and catches what's changed out in the world — and hands your AI whatever bears on the work, the moment it's needed, without you asking. The right thing, in front of the model, unprompted, with you in command of the call. So you keep moving toward what you set out to build, on top of everything you've already decided.

It plugs into Claude, Claude Code, ChatGPT and Cursor as an MCP connector. Quick to connect, in a couple of steps.

Unlimitless is open now to invited Alpha. Apply for the Beta waitlist to come in ahead of the full launch:

Alpha is invite-only · free at launch.