Skip to main content
The KIME MCP server puts the whole KIME platform inside your AI assistant. Connect Claude, ChatGPT, Cursor, Claude Code, or any client that supports the Model Context Protocol (MCP), and your assistant gets 50 tools that cover analytics, AI perception, the raw AI answers behind every metric, your prompt library, competitors, Actions, generated content, sponsored ads, and ChatGPT ad campaigns. You work with the same data the dashboard shows, but you don’t have to know where it lives. Ask a question in plain language and your assistant picks the right tools, chains them together, and explains what it found. When you’re ready to act, it can do the work for you: add and organize hundreds of prompts at once, add competitors, file and update actions, or draft ad campaign proposals for you to review.

What you can do

Track visibility and competitors

See your performance score, visibility, and share of voice against every tracked competitor, scoped to any category, tag, market, or AI model. Supported clients draw the trend as an interactive chart.

Understand AI perception

Get sentiment trends, a narrative summary of how AI describes you, the keywords driving tone, and excerpt-level scores with the source behind each one.

Read the actual AI answers

Pull the full text of what ChatGPT, Gemini, Claude, Perplexity, and the other models answered for any prompt, so your assistant can quote, compare, and fact-check them.

Research sources and citations

Rank the domains and pages AI models cite in your category, with citation counts and sentiment, to build outreach and content lists.

Build and organize your prompt library

Add up to 500 prompts in one request, create categories and tags, assign them in bulk, archive and restore prompts, and watch new runs finish.

Grow your competitor set

Add up to 500 competitors at once by name and domain. Duplicates are skipped automatically.

Work your action plan

Browse the actions KIME prioritizes for you, create new ones, change status, priority and assignee, and comment as work moves forward.

Review generated content

Read the articles, emails, and social posts KIME’s agents write for your actions. Export articles as JSON, Markdown, HTML, or WordPress (WXR).

Monitor ChatGPT ads

See which of your prompts serve sponsored ads, how coverage trends week to week, and which advertisers and creatives show up.

Plan and measure ad campaigns

Find where paid placement is worth buying, have KIME draft OpenAI Ads campaign proposals, and read live results from your connected Ads Manager account.

Check AI website health

Read your latest AI Website Health results: overall and category scores, domain and page checks, and the pages that were crawled.

Use it anywhere you work

Work from a chat app, your IDE, or your own agent, and combine KIME with the other MCP servers you use in one conversation.
For end-to-end workflows and the full list of tools, see MCP use cases.

Setup guides

1

Open the Claude Desktop app

Launch Claude on your desktop.
2

Open Connectors

Go to Customize → Connectors.
3

Add a custom connector

Click the + sign to add a connector, then select Add custom connector.
4

Enter the KIME URL

Call it KIME and enter this URL: https://mcp.kime.ai/mcp
5

Connect

Press Add, and then Connect.
The first time you ask Claude about KIME data, it will prompt you to securely sign in.
For full setup steps and troubleshooting across every client, see Connect AI assistants.

Example questions

Ask in plain language. You don’t need to name tools or IDs: your assistant looks up workspaces, brands, categories, and tags on its own. Here are some prompts to get you started. Visibility and performance
  • “How has our visibility changed over the last 30 days?”
  • “Compare our visibility in ChatGPT versus Gemini this month.”
  • “How are we performing in the [category] category in Germany?”
  • “Give me a one-paragraph performance summary I can paste into our weekly update.”
Competitors
  • “Which competitors have the highest share of voice?”
  • “Where do competitors show up but we don’t?”
  • “Add these 20 competitors with their domains and start tracking them.”
AI perception
  • “Summarize how AI models currently describe us.”
  • “Break our sentiment down by factor and show the strongest negative excerpts.”
  • “Which keywords drive negative sentiment about [competitor]?”
AI answers
  • “Show me what ChatGPT answered for our top pricing prompts last week.”
  • “Compare the AI answers for [prompt] against our product specs and flag anything wrong.”
  • “Who ranks first for [prompt], and what does the answer say about us?”
Sources and citations
  • “Which domains get cited most when our brand comes up?”
  • “Which cited pages are negative toward us? Make an outreach list.”
Prompt library
  • “We’re launching in Denmark. Set up branded and comparison prompts there and tell me when the first runs finish.”
  • “Create a ‘Pricing’ category and move every pricing-related prompt into it.”
  • “Tag all prompts that mention integrations with ‘integrations’.”
  • “Archive every prompt tagged ‘Q3 launch’.”
Actions and content
  • “What should we work on next to improve our AI visibility?”
  • “Create an action to publish a comparison page versus [competitor] and mark it high priority.”
  • “Is the article for the [topic] action ready? Show me the draft as Markdown.”
Ads and campaigns
  • “Which of our prompts serve ChatGPT ads, and who are the top advertisers?”
  • “Show me the creatives [advertiser] is running and the markets where they appear.”
  • “Where should we buy ads to catch up with competitors? Draft campaign proposals for the top 10 prompts.”
  • “How did our live ChatGPT campaigns perform over the last 30 days?”
Website health
  • “How ready is our site for AI crawlers, and what should we fix first?”

Interactive charts

In clients that support MCP Apps, such as Claude, ChatGPT, and VS Code, visibility and share-of-voice results render as an interactive KIME chart directly in the conversation. The chart follows your client’s light or dark mode. Clients without MCP Apps support receive the same data as text, and your assistant summarizes it.

Authentication

KIME uses OAuth, so there’s nothing to paste or store. The first time your assistant uses a KIME tool, it opens a secure sign-in window where you log in and approve access. After that, your session persists across conversations. On the consent screen, you see which application is requesting access and what it will be able to do, and you choose the organizations it can work in. Pick at least one. Agencies see their managed client organizations in the same list, marked Managed client, so one connection can cover every account you work on. Access always respects your KIME account: an assistant can only see the organizations and workspaces you can, and it refuses workspaces in organizations you didn’t select. To add an organization later, re-authorize the connector in your client and select the organization on the consent screen.

Read and write access

Most tools are read-only. Analytics, perception, AI answers, sources, prompts, actions, content, ads, campaigns, and website health all read existing data without changing anything. Write tools make real changes in KIME. Each one needs a specific permission, which you approve when you connect: A few things stay in the app on purpose:
  • You can’t edit a prompt’s text or permanently delete prompts. Archive a prompt instead, and restore it any time.
  • Campaign proposals are drafts. Nothing is published to OpenAI Ads and nothing is spent until you accept and publish a proposal in Ads Manager.
  • Actions are updated one at a time.
To make changes, you must be a member of the organization and select it when you connect. Write tools are flagged as such, so your client asks for confirmation before anything runs, and changes always stay within your existing KIME access.
The MCP server is the only way to make changes from outside the app. The KIME API is read-only.

Supported platforms

The KIME MCP server works with any client that supports the Model Context Protocol, including Claude, ChatGPT, Cursor, Claude Code, VS Code (GitHub Copilot), and Windsurf.

Requirements

A KIME account with at least one active workspace, and an MCP-compatible AI tool.