# IndexMachine

> Check your own sites' pages on Google and request indexing, plus IndexNow, from your AI assistant.

- name: co.indexmachine/mcp
- version: 0.1.0
- connection class: R1 (autonomous)
- trust: 100/100
- quality: 92/100 (strong)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/co.indexmachine%2Fmcp
- tools: https://api.protogrid.dev/v1/servers/co.indexmachine%2Fmcp/tools

**An agent can connect on its own if it already holds `api_key`.** Remote endpoint behind a static API key or header. The registry never sees the value; the agent fills the placeholder.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "indexmachine": {
      "type": "http",
      "url": "https://mcp.indexmachine.co/mcp",
      "headers": {
        "Authorization": "Bearer ${api_key}"
      }
    }
  }
}
```

Placeholders to fill: `${api_key}`.

## Trust

- hygiene: 100
- liveness: 100
- freshness: 100
- provenance: 100
- drivers: +repository +namespace-matches-repo +dns-namespace +website +reachable +uptime-100% +updated-7d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 92/100 (strong)

### protocol: 88 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- warn `protocol.list_ttl`: tools/list declares ttlMs 0, so clients re-fetch it on every use

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 93 (weight 30)

- pass `tools.descriptions`: every tool has a description
- pass `tools.description_length`: descriptions are concise
- pass `tools.schemas`: every tool has an object input schema
- pass `tools.annotations`: 10 of 10 tools declare readOnlyHint or destructiveHint
- warn `tools.directory_hints`: 3 of 10 tools declare all four hints; missing: account_status (idempotentHint); list_sites (idempotentHint); site_summary (idempotentHint); …
- pass `tools.token_cost`: about 2,136 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 100 (weight 15)

- pass `stability.changes`: no tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: deps.dev does not know indexmachine-mcp@0.1.0

- tools: 10, about 2,136 tokens to load them all
- tool set hash: 2c7fd40df7d41744441519885f3e370e
- badge: [![protogrid quality](https://protogrid.dev/badge/co.indexmachine/mcp.svg)](https://protogrid.dev/servers/co.indexmachine/mcp)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

_No tool definition changes recorded._

## Dependencies

- npm indexmachine-mcp@0.1.0: unknown

Dependency graphs from deps.dev; advisories from OSV.dev, including the GitHub Advisory Database and the PyPI Advisory Database; all CC BY 4.0.

## Remotes

- https://mcp.indexmachine.co/mcp (streamable-http, auth api_key ${api_key}, reachable true, uptime30d 1)

## Packages

- npm indexmachine-mcp@0.1.0 secrets: INDEXMACHINE_API_KEY

## Tools (10)

- `account_status`: The plan, the key's permissions and hourly request budget, whether the user's IndexMachine Chrome extension is online (it performs every Request indexing click), and today's requests per site against Google's daily quota of about 10 per site.
- `cancel_indexing_request`: Takes an open indexing request out of the queue, so the extension never performs it. The page stays in IndexMachine. A closed request cannot be canceled.
- `check_url`: Whether a URL of one of the user's sites is indexed on Google, and as of when: the stored Search Console inspection, its open indexing request with the ETA, and with live: true a fresh URL Inspection from Google, which uses 1 of 50 live checks per site a day. indexed is null, not known, when the URL was never inspected. When Google refuses a live check, the stored state comes back with a warning.
- `get_indexing_request`: One indexing request: its status, queue position and ETA in eta_text while open, and what Search Console answered once the extension performed it. A closed request with status requested means the click happened; whether Google indexed the page is check_url's answer.
- `list_indexing_requests`: Indexing requests newest first, for one site or all of them: open ones (preparing, queued, served) with their position and ETA, or closed ones with their outcome. Closed requests are kept 90 days.
- `list_pages`: A site's pages by path, with their status, Google's coverage state and the last crawl and request. status not_indexed is every page that is neither indexed nor excluded; q matches part of the URL. A status comes from the last stored inspection; check_url has a single URL in full.
- `list_sites`: Every Search Console property connected to IndexMachine, with its page counts (indexed, awaiting, rejected, needs_fix, excluded), whether Google indexing is on, whether a paid plan covers it (billable, which Request indexing needs) and whether IndexNow is ready.
- `request_indexing`: Queues a Request indexing click for a URL of one of the user's sites, also one that is not a page of the site yet. The user's IndexMachine Chrome extension performs it in their own Search Console when it next checks in, within about 10 requests per site per day; the answer has the queue position and the ETA in eta_text, which is for the click and not for indexing. It previews first, and queues nothing for a page that is already indexed unless force is true, a re-index. A re-index, a URL that is not in the sitemap, or the site's last request of the day waits for the user's yes. A client on MCP protocol 2026-07-28 that supports elicitation gets a question to confirm, and the answer does not pass through the model. This remote server keeps no session, so a client on an earlier protocol version is not asked: the tool refuses first with a confirm token, and the call with that token stands for the user's yes.
- `site_summary`: One site in full: page counts, today's Request indexing quota (about 10 a day, reset at midnight Pacific), the median days from a first request to indexing, and up to 30 days of counts.
- `submit_indexnow`: Notifies Bing, Yandex and the other IndexNow engines of up to 100 URLs of one site at once; it does not reach Google, and each engine decides when it crawls. It needs the site's IndexNow key file in place (indexnow_ready in list_sites) and takes 20 submissions per site a day. More than 10 URLs at once wait for the user's yes. A client on MCP protocol 2026-07-28 that supports elicitation gets a question to confirm, and the answer does not pass through the model. This remote server keeps no session, so a client on an earlier protocol version is not asked: the tool refuses first with a confirm token, and the call with that token stands for the user's yes.