Think inside your AI world.
PagerDuty through Unl
PagerDuty shows the incident and who's on the schedule. Whether it earns a 2am page is a threshold you set - and 'it's open' isn't the same as 'wake someone'.
PagerDuty through Unl weighs incident and on-call state against the paging threshold the on-call manager has ratified, so 'does this earn a 2am page?' comes back verdict-shaped, naming the unmet condition.
What PagerDuty holds
What PagerDuty through Unl pulls, unedited:
- Open and resolved incidents by service
- The service record an incident belongs to
- On-call schedules and who's covering now
- Incident timelines and status history
What the naked read gives you
PagerDuty through its MCP server returns the incident, the service it's tied to, the schedule and who's on it right now - the full, accurate picture of what's open and who's covering. That's the real API, not a summary. What it doesn't carry is the reasoning for whether this particular incident is worth interrupting someone's night: PagerDuty records severity as a field you set once, not as a live judgement about whether this specific failure, at this specific hour, clears the bar the team agreed on. The incident is open. Whether it earns the page is decided elsewhere.
The frame judges the data it is given; it does not verify the source’s accuracy.
What changes when PagerDuty is read measured
The on-call manager ratified in Unl that an incident earns an out-of-hours page only when it's tied to a Tier-1 service and open past the fifteen-minute auto-resolve window - because low-severity noise had eroded trust in the pager.
'Does this incident earn a 2am page?' returns, measured: yes - it's a Tier-1 service and it's been open twenty-two minutes, past the window the team agreed. PagerDuty supplied the incident and the service tier; your ratified rule supplied the window that turns 'open' into 'wake someone'.
And back again
Every page-worthy verdict feeds back into Unl as a record: the incident, the tier, the window it crossed, and the call that followed. Over weeks that log becomes the on-call manager's own evidence for whether the fifteen-minute window is still right - not a guess revisited from scratch each time an incident opens, but a rule tested against what actually happened.
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
Can AI tell me if an incident is worth waking someone for?
Not alone. PagerDuty through Unl reads the incident, its service and the on-call schedule; whether it earns a page depends on the tier and window you ratified in Unl. Change the window and the same open incident returns a different verdict.
How do I connect PagerDuty to Claude?
PagerDuty ships an official MCP server, detailed at github.com/PagerDuty/pagerduty-mcp-server. Run it with a user API token in read-only mode by default; Unl supplies the paging window the incident gets checked against.
Does Unl page anyone or resolve the incident?
No. PagerDuty's MCP server ships in read-only mode by default, and that's the mode Unl uses - incidents and schedules come back as data. Paging, acknowledging or resolving would need write tools enabled separately, outside Unl's connection.
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.