Cloudchipr MCP Server

Overview

You can connect your AI assistant to Cloudchipr using the Cloudchipr MCP (Model Context Protocol) server. Once connected, your assistant can access your organization's billing data, savings opportunities, live resources, and Kubernetes clusters, and answer questions like "How much could we save this month?" or "Show me idle EC2 instances in us-east-1" using your real data.

https://mcp.cloudchipr.com/mcp

Authentication

OAuth 2.0
API key

The assistant only ever reads your data. It discovers the valid accounts, dimensions, and values for your organization first, so its answers always match what you see in the Cloudchipr app.

Connect Your AI Assistant

Any MCP client that supports remote servers can connect, including Claude, ChatGPT, Cursor, and Claude Code. You can authenticate with OAuth (sign in with your Cloudchipr credentials) or with an API token.

In a GUI client (Claude, ChatGPT, Cursor)

Open your AI assistant's connector or integration settings.
Add a remote MCP server and paste https://mcp.cloudchipr.com/mcp into the URL field.
Sign in with your Cloudchipr credentials when the OAuth prompt appears.
Enable Cloudchipr in your chat and start asking questions about your cloud costs and resources.

In a CLI client (Claude Code, Codex)

Choose the method that matches how you want to authenticate.

OAuth

You will be prompted to sign in with your Cloudchipr credentials on first use.

claude mcp add --transport http cloudchipr https://mcp.cloudchipr.com/mcp

codex mcp add cloudchipr --url https://mcp.cloudchipr.com/mcp

API Token

Generate a key with Create API Key in your Cloudchipr connection settings (or paste an existing one), then replace <YOUR_C8R_PROD_TOKEN> in the command below.

claude mcp add --transport http c8r-prod https://mcp.cloudchipr.com/mcp \
  --header "Authorization: Bearer <YOUR_C8R_PROD_TOKEN>" \
  -s local

export C8R_PROD_TOKEN=<YOUR_C8R_PROD_TOKEN>
codex mcp add c8r-prod --url https://mcp.cloudchipr.com/mcp \
  --bearer-token-env-var C8R_PROD_TOKEN

Available Tools

The server exposes 25 tools across five categories. You do not call these directly: you ask your assistant a question in plain language, and it picks the right tools and discovers the valid values for your organization on its own. The example queries below show the kinds of questions each tool answers.

Cloud Accounts and Organizations

List your connected cloud accounts with total costs, underutilized costs, and amounts saved with Cloudchipr, view full details for a single account including missing permissions, and see your AWS Organizations, GCP Organizations, and Azure Tenants along with their Savings Plans and other commitments.

Tool	What it does	Example queries
`get_cloud_accounts`	Lists connected accounts with cost, waste, and savings totals.	"What cloud accounts do we have connected?" "Which account is spending the most this month?" "Show me all AWS accounts with read-write access."
`get_cloud_account`	Full detail for a single account, including missing permissions and who connected it.	"Show me the details for our production AWS account." "What permissions is the staging account missing?"
`get_provider_organizations`	Lists AWS Organizations, GCP Organizations, and Azure Tenants, with their Savings Plans and commitments.	"What AWS Organizations do we manage?" "Which accounts belong to each of our organizations?" "Do any of our organizations have a Savings Plan?"

Billing Explorer

Query your cloud spend across AWS, GCP, Azure, Kubernetes, OpenAI, Anthropic, Datadog, your custom data sources, and more. Costs can be grouped by service, account, region, tag, or dimension, filtered by any criteria, and viewed with trends and month-end forecasts over any time period. Your assistant discovers the valid values for your organization first, so results always match what you see in the app.

Tool	What it does	Example queries
`get_billing_data`	Aggregates spend by time, service, account, region, tag, or dimension, with trends and month-end forecasts.	"How much did we spend on AWS last month?" "Break down this quarter's spend by service." "What's our month-end cost forecast for GCP?" "Show daily Azure spend for the past 30 days, grouped by region."
`get_billing_providers`	Lists the providers and data sources that have billing data.	"Which cloud providers and data sources do we have billing data for?"
`get_custom_sources`	Lists your custom and third-party data sources.	"What custom data sources have we connected?"
`get_billing_groupings`, `get_billing_filters`, `get_billing_filter_values`	Discovery tools your assistant uses automatically to find the valid dimensions, filters, and values before a query. You can also ask directly.	"What can I break my AWS costs down by?" "What can I filter my GCP spend on?" "List all the regions in our AWS bill."
`validate_billing_query`	Checks a complex cost query before it runs, so results are right the first time. Used automatically behind the scenes.	(runs automatically)

Savings Opportunities

Get the organization-wide savings summary, browse opportunities sorted by monthly savings with filters for provider, region, savings amount, implementation effort, and tags, view every recommendation for a single resource, and access your saved opportunity views.

Tool	What it does	Example queries
`get_savings_summary`	The headline potential-savings figure across the whole organization.	"How much could we save this month?" "What's our total potential savings across all clouds?"
`get_provider_total_saved_amount`	The amount already saved with Cloudchipr for one provider.	"How much have we saved on AWS with Cloudchipr?"
`filter_opportunities`	Browses opportunities sorted by monthly savings, with filters for provider, region, savings amount, effort, and tags.	"Show me our top savings opportunities." "What are the highest-savings opportunities in us-east-1?" "List low-effort opportunities worth more than $100 a month."
`get_opportunity`	Every recommendation recorded for a single resource, including alternatives.	"What are all the recommendations for instance i-0a5ff2749d14ec42c?"
`get_opportunity_filters`	Discovers the filters available for opportunities. Used automatically when you ask a filtered question.	"What can I filter savings opportunities by?"
`list_opportunity_views`	Lists your saved opportunity views.	"What opportunity views do I have saved?"
`get_opportunity_view`	Opens a saved view by name to see its filters and columns.	"What's in my Quick Wins view?"

Live Resources

List live resources of any supported type (EC2, EBS, RDS, S3, VM, Disk, SQL, GKE, and more) across your accounts, filtered by region, state, utilization, or tags, with per-service totals for monthly spend and estimated waste. Drill into a single resource for full details and utilization metrics such as CPU, memory, and network. Every resource includes a direct link to its cloud console page.

Tool	What it does	Example queries
`filter_live_resources`	Lists resources of one type across your accounts, with per-service spend and estimated waste.	"Show me idle EC2 instances in us-east-1." "List all unattached EBS volumes across our accounts." "Which S3 buckets cost the most per month?"
`get_live_resource`	Full detail for a single resource, including its cloud-console link.	"Give me the full details for instance i-0a5ff2749d14ec42c."
`get_resource_metrics`	Utilization metrics (CPU, memory, network, and more) for a single resource.	"What's the CPU and memory utilization on this RDS instance over the last 7 days?" "Is i-0a5ff2749d14ec42c actually being used?"
`get_resource_tag_keys`, `get_resource_tag_values`	Discover the tag keys and values present on your resources, so tag filters match reality. Used automatically when you filter by tag.	"What tags are on our EC2 instances?" "What values does the Environment tag have?"

Kubernetes

List your Kubernetes clusters with total, idle, and reserved monthly cost alongside CPU and memory utilization, then drill into any level (nodes, namespaces, deployments, statefulsets, daemonsets, jobs, pods) to find where the spend and waste sit. View per-resource utilization, comparing actual usage against requests and limits, to right-size your workloads.

Tool	What it does	Example queries
`get_k8s_clusters`	Lists your clusters with total, idle, and reserved cost and CPU/memory utilization.	"List our Kubernetes clusters and their idle cost." "Which cluster has the most wasted spend?"
`filter_k8s_resources`	Lists objects at one level (nodes, namespaces, deployments, statefulsets, daemonsets, jobs, pods) with counts and spend.	"Show the most expensive namespaces in our prod cluster." "How many pods are running in this cluster, and what do they cost?"
`get_k8s_resource_utilization`	Per-resource CPU and memory usage versus requests and limits, the right-sizing signal.	"Compare CPU usage against requests for this deployment." "Is this node over-provisioned?"

Have specific questions? No problem at all! Please join us in Slack and drop a message in the #General channel or dm to @Ashot or @Suren.

We are fast in replies! :)

No Slack? No worries! Email us at [email protected]. We'll respond in a few hours.