Connect what helps. Nothing is required.
Connections improve the evidence. They never block a first result. Google Search Console, AI traffic from server logs, an MCP server, a CLI, an API and exports.
Evidence that improves the read, never required for a first result
One API key, minted on the Account page
Take the data anywhere
What does Inciteful connect to? In plain words.
For each client, Inciteful connects to Google Search Console and to a source of AI traffic from the client’s own server logs. For your own tools there is an MCP server, a command-line tool and a versioned HTTP API, plus exports in CSV, Markdown, HTML and JSON-LD.
Nothing is required. Connections improve the evidence, and they never block a first result.
Evidence from their side. Not inference.
Which assistants actually fetch their pages.
AI traffic reads the client’s own server logs and keeps three kinds of evidence apart. In beta.
- Retrieval fetches: an assistant pulled a page to answer someone
- Indexing crawls: a training or indexing bot read it
- Referrals: a browser arrived from an assistant
- A fetch is verified when the address belongs to the AI vendor
Search Console, read-only.
One permission, connected per client through Google sign-in. It powers Rank vs AI.
- “View Search Console data for your verified sites”
- Nothing is written back
Your agents can read it too. With one API key.
An MCP server, a CLI and an API.
So your own AI agents and scripts can use a client’s measurements directly.
- CLI: login, list, get, export, sync, health, prospect
- A versioned HTTP API over the same engine as the app
- Keys are shown once and stored hashed
So a client’s own AI tool can read its measurements and report back. A team member’s key cannot push changes.
- CSV
- Every table, and the all answers log.
- Markdown and HTML
- For documents and handovers.
- JSON-LD
- Structured data, ready to place.
What an API key can and cannot do.
- What you need
- An API key, created on the Account page. It is shown once and stored hashed.
- Where it works
- In AI applications that support MCP servers, from the command line, and over HTTP. The connect command is in the docs.
- What a key can do
- Let a client’s own AI tool read that client’s measurements and report back, through the MCP server, the CLI and the API. A team member’s key cannot push changes.
- What it never does
- It does not publish to a client’s site. Any web page the Agent fetches is treated as data, never as an instruction.
- What Google sees
- One read-only permission, per client: “View Search Console data for your verified sites”. Nothing is written back.
How client data is stored, exported and deleted is on the trust page.
Questions about integrations, answered plainly.
What does Inciteful connect to?
Per client: Google Search Console and an AI traffic source from the client’s server logs. Account-wide: an MCP server, a CLI and an HTTP API, each used with an API key.
Is any connection required?
No. Connections improve the evidence and never block a first result.
What Google permission is needed?
One read-only permission: “View Search Console data for your verified sites”. Nothing is written back.
What is the MCP server for?
So a client’s own AI tool can work with that client’s measurements directly: read the questions, tracking summaries, rival and source detail, sentiment and Site Health, and report back. Nothing reaches the client’s site except through that tool.
Is there an API and a command-line tool?
Yes. A versioned HTTP API runs over the same engine as the app, and the CLI has login, list, get, export, sync, health and prospect commands. Keys are shown once and stored hashed.
What can I export?
CSV, including the all answers log, plus Markdown, HTML and JSON-LD. You can also export all account data as JSON from the Account page.
Does Inciteful publish to WordPress or Shopify?
No. Inciteful does not publish pages or connect to a CMS. Site Health fixes are applied by your own team or your own AI coding tool.
