website-auditor-mcp

# Website Auditor MCP

An [MCP](https://modelcontextprotocol.io) server for **[website-auditor.io](https://website-auditor.io)** —
AI-visibility (GEO) and site-audit tools you can call from any MCP client
(Claude Desktop, Claude Code, Cursor, Codex, and other agents).

Ask an agent *"does ChatGPT recommend my business?"*, *"what's my AI-visibility
score?"*, *"audit example.com"*, or *"how do I stack up against my competitors?"*
and it answers with real data — an overall AI-visibility score (0–100), a
per-engine breakdown across **ChatGPT, Perplexity, Claude and Gemini**, a full
site audit (SEO, security, performance), competitor comparisons, and ongoing
monitoring.

The server is a thin, authenticated wrapper over the Website Auditor API — the
audit engine, AI-visibility scoring and monitoring live in that service; this
server just makes them available to agents.

---

## Tools

| Tool | Tier | What it does |
|---|---|---|
| `get_sample_audit` | **Free — no key** | A complete sample report for `example.com` in the exact shape a real audit returns. Needs no API key, no subscription and no network. Try this first to see what you'd be buying. |
| `get_ai_visibility` | **Pro** | Current AI-visibility score (0–100) + per-engine breakdown (ChatGPT, Perplexity, Claude, Gemini) + the top competitor appearing in your place. Pro subscribers also get `trend`: 7- and 30-day score movement from stored snapshot history for the question this audit asked, ending at its snapshot and compared only between measured snapshots that asked it (null, with a note, when the audit stored no measured snapshot). |
| `run_audit` | **Pro**, rate-limited | Full one-time audit → category scores (AI visibility, SEO, security, performance) + top issues + a shareable report URL. |
| `get_changes` | **Pro** | What changed in the AI-visibility score — the change in the overall score and the per-engine score changes, for engines measured both times; `competitor_changes`, `new_issues` and `resolved_issues` are always empty. Only between two snapshots that asked the same question (the same business name, market and queries), read from the domain's series: its measured weekly re-audits and any audit that asked the same question as the newest recorded one, while it has no gap longer than four weeks and a day from the newest weekly re-audit that recorded its question, through each later snapshot of the series (weekly re-audits included, recorded or not), to the newest measured snapshot. When there is no such pair it says why instead of a number, with the date the series re-baselined when it did. Requires the domain to be tracked. |
| `compare_competitors` | **Pro** | Head-to-head AI-visibility ranking against named competitor domains + where each appears that you don't. Quota-aware: caps the audit fan-out to your remaining daily quota, reuses recent cached audits, and reports any competitors it had to skip rather than dropping them silently. |
| `track_site` | **Pro** | Start (or stop) weekly monitoring of a site's AI visibility. Establishes the history `get_changes` reads from. |
| `untrack_site` | **Pro** | Stop monitoring a site and free up a monitoring slot. Idempotent. |
| `list_tracked_sites` | **Pro** | List the sites you're monitoring, with cadence, active state, and slots used/remaining. |
| `get_monitoring_status` | **Pro** | A glanceable dashboard across all tracked sites — latest score of each site's series (the series `get_changes` reads), the date of each one's last scheduled run and when the next runs, and the most recent like-for-like change (or a note, or the summary, saying why there is none); like `get_changes`, its `change` carries `competitor_changes`, `new_issues` and `resolved_issues`, always empty. |
| `check_upgrade_status` | Any valid key | Your own subscription standing — tier, status, period end, and what upgrading unlocks (starting Pro requires a payment method and accepting the Terms). Consumes no audit quota. |

## Prompts

Clients that support MCP prompts (Claude Desktop, claude.ai) render these as
something you can pick from a menu, so you don't have to phrase the request
yourself or know which tool to ask for. In clients that don't surface prompts
(Codex, currently), nothing is lost but the menu — ask in words and the same
tools run.

| Prompt | Needs a key? | What it does |
|---|---|---|
| **See a sample report** | **No** | Walks through a complete report for `example.com`. No arguments, no setup — one click from any install. |
| **Check my AI visibility** | Pro | Runs `get_ai_visibility` for a domain you name, then explains the score, which assistants name the business, and who is named instead. |
| **Run a full site audit** | Pro | Runs `run_audit` for a domain and summarises it by category, with the three fixes that matter most. |
| **Compare me to a competitor** | Pro | Runs `compare_competitors` for your domain against a named rival, and explains where they get named and you don't. |

Each Pro prompt falls back to `get_sample_audit` when no API key is configured,
so you always get output rather than an error.

### Naming the business (optional)

`get_ai_visibility` and `run_audit` both accept two optional arguments that
decide *what question* the AI-visibility check actually asks.
`compare_competitors` accepts `business_location` only — see below for why it
takes no name:

| Argument | Omitted | Supplied |
|---|---|---|
| `business_name` | Detected from the site, and flagged with `name_warning` when it could not be verified | Taken as fact and recorded as caller-supplied — which suppresses the warning |
| `business_location` | Detected from the site; if nothing is found the questions widen to the country, or drop the place entirely | Scopes the questions to that place |

**Supply a name only when you actually know it.** A supplied name overrides
detection and is treated as confirmed, so a guess is scored exactly as if a
human had verified it — and silences the warning that would have told you
otherwise. Leaving it out is the safer default: detection is transparent about
its own uncertainty.

The same applies to location, in the other direction. Omitting it is correct
for a national or global business and wrong for a local one, since a local
business measured without a place is measured against the wrong queries.

**On `compare_competitors`:** it takes `business_location` and applies it to
your site *and every competitor*, because a comparison asks about one market —
scoring you locally while scoring your rivals globally would rank the answers
to two different questions. It deliberately takes no `business_name`: a name
identifies one business, and forwarding yours to your competitors would score
every one of them as you. The result names the market it used, so a scoped
comparison is never mistaken for a global one.

```text
"Check AI visibility for hawaiibackroad.com,
 the business is Big Island Backroad Adventures in Hilo, HI"
```

---

## Install & configure

The server runs directly via `npx` — no clone or build required.

**Try it before you buy it.** Install with no API key at all and ask your agent
for a *sample audit* — `get_sample_audit` returns a full report for `example.com`
in the exact format a real run produces, so you can check the shape fits your
needs first.

**Pricing.** Auditing real domains needs a Website Auditor subscription at
**$10/month** — eligible new customers get a **7-day free trial** (payment
method required to start; no charge until the trial ends; customers who used
a trial in the last 12 months are billed immediately). Sign up and create an
API key at
**[api.website-auditor.io/admin_portal](https://api.website-auditor.io/admin_portal/?source=mcp)**,
then set it as `WA_API_KEY` below. There is no free API tier — a key only
functions with an active subscription.

**Claude Desktop** (`claude_desktop_config.json`), **Cursor**
(`~/.cursor/mcp.json`), and most other clients use the same `mcpServers` shape:

```jsonc
{
  "mcpServers": {
    "website-auditor": {
      "command": "npx",
      "args": ["-y", "website-auditor-mcp"],
      "env": {
        "WA_API_KEY": "wa_your_key_here"
      }
    }
  }
}
```

**Claude Code** — add it from the CLI:

```bash
claude mcp add website-auditor -e WA_API_KEY=wa_your_key_here -- npx -y website-auditor-mcp
```

**Codex** — the CLI, IDE extension and ChatGPT desktop app all read the same
`~/.codex/config.toml`, so one of these covers all three:

```bash
codex mcp add website-auditor --env WA_API_KEY=wa_your_key_here -- npx -y website-auditor-mcp
```

```toml
[mcp_servers.website-auditor]
command = "npx"
args = ["-y", "website-auditor-mcp"]

[mcp_servers.website-auditor.env]
WA_API_KEY = "wa_your_key_here"
```

Codex doesn't render MCP prompts, so the [Prompts](#prompts) above won't appear
as menu entries there — ask in words instead (*"show me a sample audit"*).

Or install the packaged **Codex plugin**, which bundles the same server and
restores the prompts as skills (which Codex does render):

```bash
codex plugin marketplace add SpikeyCoder/website-auditor-mcp
codex plugin add website-auditor@spikeycoder
```

**Cursor** — one-click install, or put the same `mcpServers` JSON as above in
`~/.cursor/mcp.json`:

[![Install in Cursor](https://cursor.com/deeplink/mcp-install-dark.svg)](https://cursor.com/install-mcp?name=website-auditor&config=eyJjb21tYW5kIjoibnB4IiwiYXJncyI6WyIteSIsIndlYnNpdGUtYXVkaXRvci1tY3AiXSwiZW52Ijp7IldBX0FQSV9LRVkiOiIifX0%3D)

The one-click config arrives with an empty `WA_API_KEY` — the sample report
works as-is; fill the key in under **Settings → MCP** to audit real domains.
There is also a packaged **Cursor plugin** (this server plus the prompts as
skills) in [`cursor-plugin/`](cursor-plugin/); its Cursor Marketplace listing
is pending — status in [docs/CURSOR-PLUGIN.md](docs/CURSOR-PLUGIN.md).

Restart the client and the tools appear.

### Getting an API key

`WA_API_KEY` is a per-user key (it starts with `wa_`) minted from a