Costory MCP & Feedback
You can now connect Costory to your favorite AI tools like Claude, Cursor, and VS Code via the Model Context Protocol (MCP).
What’s new?-
Chat with your data: Ask "Which service cost us the most this week?" and get an answer instantly.
-
Debug costs in your editor: View infrastructure spending right next to your code in VS Code or Cursor.
-
Instant support: Ask your AI assistant how to use any Costory feature: it has our full documentation on hand.
Be a beta tester
We’re looking for a few people to help us stress-test this integration. Give it a spin, find some savings, and tell us what you think.
Set it up in minutes:
AWS CloudWatch Metrics & Teams API 📊
You can now import custom usage metrics directly from AWS CloudWatch into Costory. This has been a big request from teams looking to bridge the gap between their infrastructure signals and their billing data.
By creating an AWS CloudWatch integration and pasting your Insights query (for things like S3 BucketSizeBytes, HTTP request counts, or SQS queue depth), Costory will keep these metrics in sync automatically. This makes it easier than ever to build unit economics—like cost per request or cost per GB stored—using real data.
What you can do now:
-
Sync metrics like CPU utilization or S3 storage directly to your dashboard
-
Calculate unit economics (cost per active user, cost per request) without manual exports
-
Keep your infrastructure signals alongside your billing data in one view
Check out our new AWS CloudWatch setup guide for step-by-step instructions and example queries.
Small improvements 🎨- Teams Usage Public API: New endpoints are now live. You can find the full documentation at app-api.costory.io/docs/
Billing Setup with Terraform & MCP Enhancements
Setting up your billing data is now smoother and more secure. You can now configure billing data sources for AWS, Azure, GCP, Cursor, Anthropic, and others using a fully Terraform-based workflow. This means more automation and less manual setup for your cloud cost tracking.
Improved MCP ExperienceWe've polished the AI Assistant (MCP) experience to make your interactions faster and more intuitive. You can also reach costory docs from the MCP.
Small Improvements 🛠️-
Costory is at KubeCon this week! If you're there, come say hi.
-
Next release is scheduled for April 1st.
Introducing Teams and MCP (beta)
A single cost view rarely works for an entire organization. Your backend squad needs Contracted Cost scoped to their services, while finance needs Billed Cost across everything. Teams let you configure those defaults once so each group lands on the right numbers without manual filtering.
What’s new:
-
Configure custom default views for different departments
-
Scope metrics (Contracted vs. Billed) to match team-specific needs
-
Self-service navigation so teams land on the right numbers automatically
Our MCP server is officially in beta! We're building a more seamless way to interact with your data through AI, and we'd love for you to help us shape it.
-
Connect your AI assistants directly to Costory data
-
Provide feedback during the beta to influence the final version
Check out our documentation on AI Assistant (MCP) to get started.
Events Page & New Budget Tools
Tracking deployments, incidents, and infra changes shouldn't feel like detective work. We've added a dedicated Events page so you can see exactly what happened and when, with full filtering and tagging support.
What's new:
-
Centralized View: Browse all events with metadata, tags, and provider info at a glance.
-
Instant Insights: Use the inline explorer modal to see the cost impact of any single event without leaving the page.
-
Save as Context: Jump directly from an event into the Advanced Explorer with filters pre-filled to drill into the "why" behind cost spikes.
Setting up budgets just got a lot less tedious. We've added new filling options to help you configure your spend targets exactly how you want them.
-
Flexible Allocations: Set monthly targets manually or distribute a total amount evenly across the year.
-
Smart Baselines: Use your historical spend as a starting point, then tweak as needed (no more guessing from scratch).