> ## Documentation Index
> Fetch the complete documentation index at: https://elarislabs-87.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Elaris Visibility in Claude and ChatGPT

> Connect Elaris Visibility to Claude, ChatGPT or another assistant, and ask about your numbers in plain words. The connection is read-only and reads only the workspaces you choose.

## What you can ask

* “Which topics are we losing to competitors in Qatar this month?”
* “What does ChatGPT say about us when people ask for sofa shops?”
* “Who is asking for a recommendation on Reddit this week, and which ones are worth a reply?”
* “What should we fix first, and why?”

The assistant looks up the same numbers you see in the app, with how many answers each rests on, and links to the page that shows more.

## Connect Claude

1. In Claude, open **Settings → Connectors** and choose **Add custom connector**.
2. Name it Elaris Visibility and paste `https://visibility.elarislabs.ai/mcp`.
3. Press **Connect**. Sign in to Elaris Visibility if asked, choose the workspaces Claude can read, and press **Allow read-only access**.

> [!NOTE]
>
> Custom connectors are available on Claude's paid plans. On a Team or Enterprise plan, an owner may need to add the connector for the organisation first.

## Connect ChatGPT

1. In ChatGPT's settings, under **Connectors**, turn on developer mode (in the advanced settings).
2. Create a connector with the address `https://visibility.elarislabs.ai/mcp` and OAuth sign-in.
3. Sign in to Elaris Visibility, choose the workspaces and press Allow.

> [!NOTE]
>
> Which ChatGPT plans can add custom connectors is up to OpenAI and changes from time to time.

## Claude Code, Cursor and other tools

Claude Code signs in the same way: run `claude mcp add --transport http elaris-visibility https://visibility.elarislabs.ai/mcp`, then type `/mcp` in Claude Code and sign in.

For tools that send a header instead of signing in, create a personal token in **Settings → Integrations → AI assistants** and send it as `Authorization: Bearer <token>` to the same address. The token is shown once; keep it private, as anyone who has it can read what you can.

## What it can read

| Tools | What they return |
| - | - |
| `list_workspaces` | The workspaces it can read, with each region's country, languages and engines |
| `visibility_overview` | Visibility, share of voice, citation rate, sentiment and brand injection, with their change and by engine |
| `brand_ranking` | Every brand AI names, ranked, with position and sentiment |
| `topics, topic_detail` | Topics with their demand and visibility, and one topic's gaps and prompts |
| `search_prompts, read_answers` | Find prompts, and read what each engine actually answered |
| `citations, compare_competitor` | The sites and pages AI cites, and how you compare with one competitor |
| `sentiment, accuracy` | How AI describes you, and the claims it gets wrong or out of date |
| `social_listening, social_leads` | What people post about you and your competitors, and the leads in it |
| `suggestions` | Action Center suggestions, most likely to pay off first |
| `setup, bot_readability, ai_traffic` | How the region is set up, whether AI can read your site, and AI visits to it |

## How access works

* **Read-only.** An assistant can't change settings, run collection, mark anything or use credits.
* **As you, in the workspaces you choose.** It sees only what you can see in the app, and only the workspaces you ticked when you connected it (or every workspace you can open, for a personal token unless you limit it).
* **Yours to remove.** Every connection and token is listed in **Settings → Integrations → AI assistants**; Remove ends it at once. Connected apps get a token that lasts an hour and renews itself while you keep using it.
* **Paused with the workspace.** When a workspace's trial ends or its plan stops, assistants can't read it until it's restarted.

## Technical details

The connector is an MCP server (Streamable HTTP) at `https://visibility.elarislabs.ai/mcp`. Assistants sign people in with OAuth 2.1: dynamic client registration, the authorization code flow with PKCE, and refresh tokens; discovery starts at `/.well-known/oauth-protected-resource`. Personal tokens are sent as a bearer header. Questions? See [FAQ and support](https://elarislabs-87.mintlify.site/visibility/support.md).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.