May was about closing the loop on ops workflows. Until now, you could query databases, manage deployments, and publish content through your AI — but if something broke, you were switching back to Grafana dashboards and Prometheus queries manually. That gap is closed now.
Four new integrations this month, covering observability, container management, and project tracking. Plus some dashboard and onboarding fixes based on what early users told us wasn't working.
Here's what shipped.
TL;DR: May 2026 Update
- Prometheus — run instant and range queries against your metrics directly from your AI client.
- Grafana — list alert rules, check what's firing, manage silences, and act on it — all in one conversation.
- Portainer — list, inspect, start, stop, restart containers, and deploy stacks without opening a separate dashboard.
- GitHub Boards — manage your Project v2 boards, create items, update statuses, and archive or delete — from wherever you're working.
- Dashboard & onboarding fixes — several rough edges smoothed out based on feedback from alpha users.
✨ What's New 🚀
Your Metrics, in the Conversation
Observability has always been a context-switching problem. Something looks off in production. You open Grafana, dig through dashboards, run a Prometheus query, cross-reference the alert — then come back to your AI to discuss what you found. That's three tools and two tab switches to answer one question.
Prometheus and Grafana ship together this month because they're most useful that way.
With Prometheus, your AI can run instant queries at a specific point in time or range queries across a time interval — the same PromQL you'd write yourself, triggered in plain language. Ask it to check CPU usage over the last 30 minutes, pull error rate for a specific service, or compare latency across instances.
Grafana covers the alerting side. Your AI can list all your alert rules, pull what's currently firing (with optional label filters so you're not wading through noise), and list active silences. When you need to act — create a silence for a specific alert rule, or delete one that's expired — that's available too.
In practice: a client messages you at 9pm about degraded response times. Open your AI client, ask it to pull the relevant Prometheus range query and check Grafana for active alerts. It surfaces what's firing, you make the call, you silence the noise. No dashboard hopping. The whole triage conversation happens in one place.
Your Containers, Without the Dashboard
The Portainer integration this month is for anyone managing Docker environments who's tired of opening a UI just to check what's running.
Your AI can list all your Portainer endpoints, list stacks across all of them, and drill into containers for a specific endpoint. From there: inspect a container's full configuration, start it, stop it, restart it, or deploy a new Docker Compose stack from inline content.
That last one is worth noting. You can hand your AI a Compose file — or ask it to draft one — and deploy it directly. No copy-paste into a UI, no manual trigger. Draft, review, deploy, done.
Combined with the Prometheus and Grafana integrations above, you now have a full incident-to-resolution loop available in a single conversation: alert fires → pull metrics → inspect the relevant container → restart or redeploy → confirm it's back.
Your Project Board, Where You Already Are
GitHub Boards — specifically GitHub Projects v2 — is the last integration this month, and it's a practical one for anyone using GitHub for project tracking alongside code.
Your AI can list all Project v2 boards for a user or organisation, retrieve details for a specific board, and create new ones. Within a board: list items, create draft items or issue-linked items, update the status of any item using a human-readable status name (no field IDs required), archive items, delete them, or pull full details on a specific item.
In practice: you finish a PR, ask your AI to move the corresponding board item to "Done," and create a follow-up item for the next task — without opening GitHub's project UI. For anyone context-switching between code and project tracking all day, that's a meaningful reduction in friction.
🛠️ Fixes & Improvements
We Fixed What You Told Us Was Broken
Alpha feedback has been direct, which is exactly what we wanted. Several users flagged rough spots in the dashboard and onboarding flow — things that weren't obviously broken but added unnecessary friction when getting started.
Those fixes shipped this month. The onboarding sequence is cleaner, the dashboard surfaces what you need faster, and a handful of edge cases that confused setup have been resolved.
Tell us what to build next.
We're adding integrations every month based on what freelancers actually need. If there's a tool missing from your workflow,
drop us a line.
Try It Now
Prometheus, Grafana, Portainer, and GitHub Boards are all available today. The free tier covers everything you need to get started — no credit card required. Your first MCP server takes under 5 minutes to set up.
Create your free account →
