Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@getcited Audit example.com and summarize the top SEO issues."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Client-side pieces for connecting a coding agent to GetCited: the MCP server definition, a Claude Code plugin with its skills, a Gemini CLI extension, and a small stdio proxy for clients that cannot speak Streamable HTTP.
GetCited audits a site the way an answer engine reads it, tracks where it ranks, records which AI engines cite it, and files each measured problem as an action with the evidence behind it. It never proposes the fix: your agent has the repository open and it does not.
The service itself is closed source. Everything in this repository is the connector — transport, manifests and prompts — so you can read exactly what your agent is being told and what leaves your machine.
Create a key under Settings → API keys in the app, then pick your client.
Claude Code
claude mcp add --transport http getcited https://app.getcited.dev/api/mcp \
--header "Authorization: Bearer $SEO_API_KEY"Any client that reads a JSON config
{
"mcpServers": {
"getcited": {
"type": "http",
"url": "https://app.getcited.dev/api/mcp",
"headers": { "Authorization": "Bearer ${SEO_API_KEY}" }
}
}
}Gemini CLI
gemini extensions install https://github.com/bapierre/getcited-mcpClients without a remote transport (Claude Desktop, whose config takes a
command rather than a URL) use the stdio proxy in src/:
claude mcp add getcited --env SEO_API_KEY="vseo_..." -- npx -y getcited-mcpVariable | Required | Meaning |
| yes | The key from Settings → API keys. |
| no | Endpoint override. Defaults to |
llms-install.md is the same instructions written for an agent, so you can tell
your assistant to set this up rather than doing it yourself.
Related MCP server: Website Auditor MCP
Path | What it is |
| The stdio proxy. Forwards JSON-RPC to the HTTPS endpoint and nothing else. |
| The Claude Code plugin and its two skills. |
| The Gemini CLI extension manifest. |
| A project-scoped Claude Code server entry. |
| Setup instructions addressed to an agent. |
Sites (list_projects, add_project, remove_project, get_site_health), the
queue (list_actions, get_action, claim_action, complete_action,
dismiss_action, run_audit), rankings (add_keywords, list_keywords,
rank_history, check_rankings), AI visibility (add_geo_prompts,
geo_summary, check_geo, run_brain) and visitors (get_behavior_digest,
get_page_profile).
Each tool carries a title and the readOnlyHint or destructiveHint
annotation that matches what it does. Full documentation, including what every
number means and the rules the measurements follow:
https://getcited.dev/docs/mcp.
Listed in the official MCP Registry as dev.getcited/mcp, published under a
DNS-verified namespace.
getcited.dev/support · privacy · terms · MIT licensed.
This server is not authenticated, so none of its SEO tools are loaded. Call this for the exact steps to fix it. Retrying other tools will not help.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already mark this as read-only, idempotent, and non-destructive. The description adds context that the server is unauthenticated and that other tools are not loaded, which explains why this tool exists. It doesn't contradict annotations and adds useful context about the state, so it earns above baseline.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is three short sentences, each earning its place: the state, the action, and a caution against retrying other tools. It is front-loaded with the key information and contains no filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a tool with no parameters and no output schema, the description covers everything an agent needs: what the tool is for, when to call it, and what to expect (steps). The guidance about other tools is an extra that improves completeness. Nothing is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, and the schema is empty with 100% coverage. Per the rubric, the baseline is 4 for zero parameters, and the description adds no parameter-specific information, which is appropriate since there are none. No further explanation needed.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool provides 'exact steps to fix' the authentication issue, with a specific call to action. It distinguishes itself by being the dedicated tool for this situation, and the name reinforces the purpose. It could be slightly more explicit about returning instructions, but it's unambiguous.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly tells the agent when to use it: when the server is not authenticated. It also gives a clear exclusion: 'Retrying other tools will not help,' preventing wasted attempts on other tools. No alternatives exist (no siblings), so this guidance is complete.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0how_to_authenticateScored across 1 tool
Only one tool is exposed, so there is no possibility of confusing it with another operation. Its purpose as an authentication helper is clearly stated.
The single tool name how_to_authenticate is descriptive and readable, but there is no established pattern to evaluate since it is the only tool. It does not follow a typical verb_noun naming convention.
A single authentication helper is far too few for a server meant to provide SEO tools. The current toolset is essentially a placeholder rather than a functional suite.
The intended SEO tools are entirely absent; the only available tool tells agents how to authenticate. This leaves the server's core domain completely uncovered and any real task impossible to complete.
Audit any website for AI visibility: graded report, findings with fixes, AI crawler access check.
Site scores, audits, Search Console, keywords, competitors, authority and AI visibility tools.
Public-site SEO/AEO readiness scans with evidence scores, issues, and clear measurement scope.
Run SEO + AI-visibility (GEO) audits from Claude, Cursor & other AI clients.
