Stackpulse

Stackpulse puts the services your product depends on (out of 9,700+ official status pages: AWS, Cloudflare, Stripe, Twilio, GitHub…) on one dashboard with your own websites and API endpoints, checked every minute. This connector gives Claude that dashboard and its history, so it can answer the questions a single status page can't.

Things to ask:

• "Is anything on my stack down right now?" Every vendor and URL on your dashboards, problems first.
• "Our API got slow around 3 pm. Is it us or one of our vendors?" Your URL monitor's checks beside your vendors' problems from that time.
• "What went wrong on our stack this week?" Every incident across your vendors and URLs, with what else broke at the same time.
• "Which of our vendors was least reliable this quarter?" Time with problems by severity, incidents and recovery for each, and which vendors tend to break together.
• "How has our checkout endpoint done this month?" Uptime, response times, failed checks and its certificate.
• "Add Twilio and Postmark to our Production dashboard and watch api.example.com."

With Read & write access it can also create and rename dashboards, add or remove services, add URL monitors, run a check now, and pause or resume one. The only removal is taking services off a dashboard. Any of the 9,700+ services can be looked up too, dashboard or not.

You need a Stackpulse account on Pro ($39 a month; every new account gets a 14-day trial, no card). You sign in with Stackpulse, the connection starts read-only and asks again before it may change anything, and you can limit it to some dashboards. Disconnect any time in Settings → API & MCP.