Back to all days

Day 16 — DevOps

CI/CD Status Dashboard

“Is the build broken?” shouldn't require opening 4 tabs.

An MCP server that tracks build status, deployment history, test results, and uptime across every project on your team. One question instead of four dashboards.

Example Queries

“Is the build broken on any project right now?”
“Show me the failure rate for the mobile app over the last two weeks”
“What was the last deployment to production for the API server?”
“Which team has the most build failures?”
“Is auth-service healthy? Any recent downtime?”
“Give me the full CI/CD dashboard summary”

Tools

6 MCP Tools

ToolWhat It Does
build_statusGet recent build status across all projects or for a specific project. Shows pass/fail, duration, author, and commit info.
deploy_historyView deployment history across projects. Shows version, environment, status, and rollbacks.
pipeline_healthHealth summary for each project — success rate, average build time, last build status, and failure trends.
build_detailsDetailed breakdown of a specific build — all stages with durations, statuses, and error logs.
uptime_reportUptime and response time report for deployed services. Shows availability percentage and any incidents.
dashboard_summaryExecutive overview — total projects, success rates, deploy frequency, active incidents, and team activity.

Coverage

What the Dashboard Tracks

Build Status

Pass/fail across all projects, commit info, authors, and build times

Pipeline Health

Success rates, average duration, failure trends, and team breakdowns

Deployments

Version history, environments, rollbacks, and deploy frequency

Uptime Monitoring

Availability percentages, response times, and incident detection

Build Stages

Stage-by-stage breakdown with error logs for failed builds

Team Activity

Which teams are shipping, which are stuck, and where bottlenecks live

Why This Build

Why CI/CD Visibility?

Engineering teams live in CI/CD. But the tools that run their pipelines — GitHub Actions, CircleCI, Jenkins — show you one repo at a time. If you manage 5 services, that's 5 tabs. If you manage 20, good luck.

The real questions aren't “did this one build pass?” — they're “is anything broken right now?”, “which team has the most failures this week?”, and “when was the last production deploy for auth?”

This dashboard consolidates builds, deploys, and uptime into one queryable interface. Instead of clicking through dashboards, just ask. First DevOps build of the challenge.

Demo Data

What's Inside

8

Projects

14

Pipelines

61

Builds

13

Deployments

245

Uptime Checks

5

Teams

8 projects across 5 teams (Frontend, Backend, Mobile, Data, Platform) with 14 days of build history. Includes a flaky mobile test suite, a recent auth-service outage with rollback, a currently-running build, slow data pipeline builds, and mixed deployment environments. Edge cases: failed deploys, rollbacks, and a 90-minute brief service outage.

Setup

Try It Locally

cd day16-cicd-dashboard
npm install && npm run seed && npm run build

Then add the config to Claude Desktop and restart.

Want visibility across your entire pipeline?

If your engineering team is tab-switching between GitHub, Vercel, and Datadog just to answer “is anything broken?” — there's a better way.