# One site (`mttr_site`)

Updated October 24, 2026.

## What does mttr_site do?

One site as its latest read found it: the score, the four stages, and every fix in order with each page that fails it. Any key may make this call, on any public site.

Who may call it: Any key.

## When to use it

- You want the figures behind a site’s report.
- You want every failing page, not just a few.

## What you ask

You might say: “How is example.com doing?”

```
mttr_site({ site: "example.com" })
```

| Argument | Type | Needed | What it is |
| --- | --- | --- | --- |
| `site` | Text | Yes | The site, as a domain such as example.com. |

## What comes back

- The score and each of the four stages, with the letter grade the report card gives each.
- Every fix in order of worth, with the pages failing it.
- Its place among every ranked site, the towns its pages say it serves, and its report card to pass on.

A sample, with made-up sites and figures:

Score 75 of 100. Fix first: Show sources on guides.

| Stage | Score |
| --- | --- |
| Access | 92 |
| Understand | 71 |
| Trust | 48 |
| Act | 80 |

A score says pages are well formed. It does not say an assistant cites the site.

## Over HTTP

```
curl "https://mttr.si/api/v1/sites/example.com" \
  -H "Authorization: Bearer $MTTR_KEY"
```

## When it fails

| When | Status | What to do |
| --- | --- | --- |
| Your assistant has no key yet | 401 over MCP | Sign up with `mttr_signup`, then connect again with the key it gives. |
| mttr has no public report of the site yet | 404 | Ask for a read with `mttr_rescan`, then watch it with `mttr_job`. |

## Sources

- [Model Context Protocol: tools](https://modelcontextprotocol.io/specification/2025-06-18/server/tools), Model Context Protocol: How an assistant calls a tool.

The call in full: [mttr_site in the API reference](https://mttr.si/docs/reference#mttr_site).

## Goes with

- [Brief](https://mttr.si/docs/tools/brief.md)
- [Plan](https://mttr.si/docs/tools/plan.md)