get_metricsLABSCO SUMMARY
Field data that arrives already rated.
It answers Core Web Vitals questions from real visitors rather than a lab run, and every figure comes back with its good, improve or poor rating and the filters that produced it — so the number arrives with the context needed to act on it.
What it is
A hosted query layer over CoreDash, a real user monitoring platform for Core Web Vitals. It answers from field data collected from real visitors, segmented by page template, country, device, browser, OS and a long list of other dimensions.
What you get
- Current performance scores at p75, filtered and grouped by any dimension
- A metric over time with an improving, stable or regressing summary attached
- The distribution of a single metric across roughly 40 buckets, which is how a bimodal or long-tailed problem becomes visible
- Rated buckets on every answer — good, improve, poor — matching the Core Web Vitals thresholds
- Filters on device, country, page path, URL, browser, OS and the LCP, INP and CLS attribution element selectors
Requirements
A CoreDash account. Point your client at https://app.coredash.app/api/mcp and complete the browser OAuth sign-in, or send an API key as a bearer token for CI. Nothing installs. Data is stored in the EU, and paid plans are billed monthly.
Setup effort
One command — claude mcp add --transport http coredash https://app.coredash.app/api/mcp
USE CASES
01
Finding out whether the site is actually fast for users
p75 LCP, INP and CLS from real visitors, grouped by whichever dimension you suspect.
02
Catching a regression after a release
The metric over time with an improving, stable or regressing verdict attached, so the trend is read rather than eyeballed.
03
Explaining an average that does not match the complaints
The distribution across roughly 40 buckets shows whether traffic is bimodal or long-tailed — which the average hides.
04
Narrowing a problem to a template or a country
Filter by page path, device, country or browser and see where the poor bucket concentrates.
05
Finding the element responsible for LCP
Attribution element selectors are filter dimensions, so the slow element is isolated rather than inferred.
3 tools
Listed in project docs. Your client may show a different list.
get_timeseriesget_histogramBEFORE YOU INSTALL
What to know first
Tools you get
3 tools named in the project's documentation.
Requires
A client that can reach a remote MCP server over HTTP
Where it runs
All 5 documented clients connect to a hosted endpoint.
INSTALLATION
Connect your client
Server setup
claude mcp add
Claude Code
TERMINAL
claude mcp add --transport http coredash https://app.coredash.app/api/mcpHosted endpoint
ENDPOINT URL
https://app.coredash.app/api/mcpAdd this hosted endpoint to your MCP client.
MCP Clients
Cursor, Any MCP client
JSON
{
"mcpServers": {
"coredash": {
"url": "https://app.coredash.app/api/mcp"
}
}
}Windsurf
WINDSURF · JSON
{
"mcpServers": {
"coredash": {
"serverUrl": "https://app.coredash.app/api/mcp"
}
}
}VS Code
VS CODE · JSON
{
"servers": {
"coredash": {
"type": "http",
"url": "https://app.coredash.app/api/mcp"
}
}
}
