Think inside your AI world.
New Relic through Unl
New Relic shows the error rate and the deploy marker beside it. Whether the deploy caused the spike is your rule to apply - and 'they're close together' isn't the same as 'it's the cause'.
New Relic through Unl holds error rate and deploy markers against the attribution rule the release manager has ratified, so 'did the deploy cause this?' returns verdict-shaped, naming the condition unmet.
What New Relic holds
The real NRDB access behind New Relic through Unl:
- NRQL queries run directly against NRDB
- Dashboards and the charts inside them
- Alert policies and open or closed events
- Recent logs for the account
- Entities looked up by GUID, name or tag
What the naked read gives you
New Relic's MCP server runs NRQL directly against NRDB, lists dashboards and their charts, pulls alert policies and recent events, fetches logs, and looks up entities by GUID or tag - the platform's real querying power, exposed as tools. That's substantial, current access. What it doesn't do is attribute cause: the error rate rose and a deploy happened nearby, but whether the deploy is the reason is a rule about timing windows and blast radius that New Relic's data doesn't encode. The two events sit close together on the timeline. Whether one caused the other is decided by a rule someone wrote down.
The frame judges the data it is given; it does not verify the source’s accuracy.
What changes when New Relic is read measured
The release manager ratified in Unl that a deploy only counts as the cause of an error spike when the spike starts within ten minutes of the deploy marker and the affected service was actually touched by that deploy - because 'it happened afterwards' had been blamed on the wrong release twice.
'Did the deploy cause this?' returns, measured: no - the error spike started forty minutes after the marker, and this deploy didn't touch the affected service. New Relic supplied the error rate and the deploy marker; your ratified rule supplied the timing-and-scope rule that decides attribution.
And back again
The attribution call - deploy caused it, or didn't - gets logged back to Unl with the timing gap, the service scope and the rule it was tested against. The next error spike near a deploy doesn't start the argument over; the release manager has the prior ruling and the rule that produced it, ready to apply or amend.
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 my deploy caused the error spike?
Not from the timeline alone. New Relic through Unl reads the error rate and the deploy marker; whether that counts as cause depends on the timing-and-scope rule you ratified in Unl. Two events close together and two events forty minutes apart return different verdicts.
How do I connect New Relic to Claude?
New Relic runs a hosted MCP endpoint, documented at docs.newrelic.com/docs/agentic-ai/mcp, reachable via OAuth or an API key. Bring that endpoint into Unl, where the attribution rule for what counts as 'caused by the deploy' actually lives.
Does Unl roll back a deploy or edit an alert policy?
No. The New Relic connection is one-directional: NRQL results, alert state and logs flow into Unl, nothing flows back to New Relic. Rolling back a release or changing a policy stays a decision made in New Relic directly.
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.