Vantage MCP

@vantage show me our cloud spending breakdown by service for the last 30 days

Based on your Vantage account data, here's an illustrative breakdown: AWS EC2 accounted for $12,400 (41%), RDS for $7,200 (24%), S3 for $3,800 (13%), and other services totaling $6,600 (22%). Total spend was $30,000 across 847 resources. The largest cost increase compared to the previous period was EC2, up 18% due to additional m5.2xlarge instances in us-east-1.

This example demonstrates cost visibility across cloud providers. Requires OAuth2 connection to your Vantage workspace with read access to cost reports. Data freshness depends on your cloud provider's billing sync schedule—typically 24-48 hours behind real-time usage.

@vantage create a budget alert for our staging environment if monthly costs exceed $5000

I've created a budget alert in your Vantage workspace targeting resources tagged 'environment:staging'. You'll receive notifications via your configured channels (email, Slack) when cumulative monthly costs cross the $5,000 threshold. The alert checks hourly and includes a 10% warning threshold at $4,500. Current staging spend this month is $3,240.

This write action creates a persistent alert in Vantage that triggers outside Switchy. Requires OAuth2 scope for managing budgets and alerts. The alert relies on accurate resource tagging in your cloud accounts—untagged staging resources won't be monitored.

@vantage analyze our cost trends and recommend three specific optimizations we should prioritize this quarter

After reviewing your Vantage cost data, here are three high-impact opportunities: (1) Right-size 14 over-provisioned RDS instances in production—estimated $2,100/month savings with minimal risk. (2) Migrate infrequently accessed S3 data (340GB) to Glacier—saves $180/month. (3) Purchase a 1-year Reserved Instance commitment for your baseline EC2 load—projected $4,800 annual savings. These recommendations account for your current usage patterns and growth trajectory over the past 90 days.

This synthesis example combines Vantage's cost intelligence with AI reasoning to surface actionable recommendations. The quality depends on how comprehensively your cloud resources are connected to Vantage. Recommendations assume stable workload patterns—validate against upcoming project plans before committing to reservations.

When your AWS bill jumps 40% and you need answers fast

A 6-person engineering team wakes up to a Slack alert: last month's AWS spend is up $8,000. The Vantage MCP lets your AI agent pull cost breakdowns by service, tag, and time window without anyone leaving the standup channel. You ask "what drove the spike in us-east-1 last week" and get a structured answer in 10 seconds. This works because Vantage already aggregates your cloud billing data—the MCP just surfaces it through natural language. The trade-off: if your team doesn't already use Vantage for cost tracking, you're adding a new platform just to query it. But if Vantage is already your source of truth, this MCP turns cost forensics from a 20-minute dashboard hunt into a 2-question conversation. Worth it for teams running multi-account AWS or GCP setups where cost attribution is messy.

Pulling cost reports for finance without the spreadsheet dance

A startup's finance lead needs a breakdown of engineering spend by project for the board deck. Normally this means exporting CSVs from three cloud providers, pivoting in Excel, and reconciling tags. With the Vantage MCP, your AI agent can pull pre-aggregated cost data filtered by custom dimensions—team, environment, or cost center—and format it for slides in one prompt. The MCP handles OAuth so the agent accesses your Vantage account securely without API keys floating around. This scenario assumes you've already tagged resources consistently in Vantage; if your tagging is a mess, the MCP won't fix that. Best fit for teams that run monthly or quarterly cost reviews and want to cut the manual export-pivot-paste loop from 2 hours to 5 minutes.

Checking deployment cost impact before merge

A platform team wants to block PRs that would spin up expensive infra without approval. They write a GitHub Action that uses the Vantage MCP to query forecasted costs for the new resources in the Terraform plan. If the 30-day projection exceeds $500, the action comments on the PR with a breakdown and requests finance sign-off. This works because Vantage can model cost impact from IaC changes—the MCP makes that data available to your automation. The catch: this requires Vantage's cost forecasting feature and tight integration with your IaC workflow. It's overkill for teams under $10k/month in cloud spend, but it's a forcing function for cost discipline at scale. If you're already using Vantage and your infra changes are frequent, this turns cost governance from a post-mortem into a pre-commit check.