MCP: questions and answers

The Model Context Protocol, memory servers and connecting AI to your own systems.

  • What is an MCP server?

    An MCP server is a program that exposes capabilities to AI applications through the Model Context Protocol (MCP): tools the model can call, resources it can read and prompt templates a user can pick. The AI app (the host) connects to it through an MCP client. A server runs locally on your machine over stdio, or remotely over HTTP, and any MCP-capable app can use it.

  • What is Mem0, and is it open source or a paid platform?

    Mem0 is a memory layer for AI agents and apps: it stores facts from conversations outside the model, so an agent remembers across sessions. It comes in two forms. Mem0 Open Source is an Apache 2.0 library or Docker stack you run on your own infrastructure. Mem0 Platform is the hosted version, which as of November 2026 has a Free plan with usage limits and a Custom plan on request.

  • What is the AWS MCP Server and how do you set it up safely?

    The AWS MCP Server is a managed, remote MCP server run by AWS, generally available since May 6, 2026. It gives AI agents such as Claude Code, Cursor, Codex and Kiro a small set of tools: call any of 15,000+ AWS API operations, search and read AWS documentation, and run short Python scripts in a sandbox. It acts with your IAM permissions and costs nothing extra.

  • What is the GitHub MCP server and how should a team set it up?

    The GitHub MCP server is GitHub's own MCP server. It lets an AI tool such as Claude, Copilot, Cursor or Codex work with your repositories, issues, pull requests and Actions. A team usually connects the remote server at https://api.githubcopilot.com/mcp/ with OAuth, enables only the toolsets it needs, and uses read-only mode where the AI should not write.

  • What does the Figma MCP server give a coding agent, and which seat do you need?

    Figma's MCP server gives coding agents such as Claude Code, Cursor, Codex and VS Code a frame's design context: variables, components, layout and Code Connect mappings. The remote server at https://mcp.figma.com/mcp is recommended, on all plans. As of November 2026, read calls are limited per seat: 20 a month for View and Collab seats, and 200 to 600 a day for Dev and Full seats on paid plans.

  • How do you build a custom MCP server, for example in Python?

    Use an official MCP SDK: Python, TypeScript, C#, Go, Rust and Ruby are Tier 1. In Python you install mcp[cli], create an MCPServer, turn functions into tools with @mcp.tool(), and run it over stdio for one machine or Streamable HTTP for a team. Test it with the MCP Inspector, and add token verification before anyone else connects to it.

  • How do you connect Jira and Confluence to AI with the Atlassian MCP server?

    Atlassian runs its own remote MCP server, formerly called the Atlassian Rovo MCP server, at https://mcp.atlassian.com/v2/mcp. It connects Jira, Jira Service Management, Confluence, Bitbucket and more to MCP clients such as Claude, ChatGPT, GitHub Copilot, Cursor and Codex. Users sign in with OAuth 2.1, the AI sees only what that user may see, and org admins control domains, IP ranges and read, write and search permissions.

  • What is the Playwright MCP server, and when should a team use it?

    Playwright MCP is Microsoft's open source MCP server that lets an AI agent drive a real browser with Playwright. It works from the page's accessibility tree instead of screenshots, so a text model can use pages. It runs with npx @playwright/mcp@latest in Claude Code, Cursor, Codex and VS Code. Microsoft says it is not a security boundary, and that for coding agents a CLI with skills is often more token-efficient.

  • Is there an official Slack MCP server, and what can the AI see?

    Yes. Slack hosts its own MCP server at https://mcp.slack.com/mcp, over Streamable HTTP with user OAuth. Slack names Claude.ai, Claude Code, Cursor and Perplexity as supported clients. The AI acts on behalf of the signed-in user and can search, read and post where the granted scopes allow, which can include private channels and DMs. Workspace admins approve MCP clients and can audit activity.

  • What is the Azure MCP Server and how should a team use it?

    The Azure MCP Server is Microsoft's open source MCP server for Azure. It lets AI agents such as GitHub Copilot agent mode, Claude Code and Cursor list, query and change Azure resources across more than 45 service areas. It signs in with your Entra ID identity, so the agent can do exactly what your Azure RBAC role allows, and it has a read-only mode.

  • What is the Datadog MCP server and how do you set it up for a team?

    The Datadog MCP server is a remote MCP server run by Datadog, generally available since March 9, 2026. It lets AI agents such as Claude Code, Cursor, Codex and Copilot query logs, metrics, traces, monitors, dashboards and more. It uses the signed-in user's own permissions, needs MCP Access switched on by an admin, and only exposes the core toolset unless you ask for more.

  • What is the Filesystem MCP server and is it safe to use at work?

    The Filesystem MCP server is one of the reference servers of the Model Context Protocol project, published as @modelcontextprotocol/server-filesystem. It lets an AI tool read, write, edit, move and search files, within its active allowlist, which older clients can replace through Roots unless you enable --confine-roots. A --read-only flag removes every write tool. The maintainers call it an educational example, not a production-ready solution.

  • How do you connect Claude or ChatGPT to Salesforce with an MCP server?

    Salesforce runs its own hosted MCP servers, now generally available, for Enterprise Edition and above. An admin switches on each server under Setup, API Catalog, MCP Servers, and creates an External Client App for OAuth. Claude, ChatGPT, Cursor and other MCP clients then work with Salesforce data with exactly the permissions of the person who signed in, no more.

  • How does the Notion MCP server work for a team?

    Notion hosts its own MCP server at https://mcp.notion.com/mcp, over Streamable HTTP. Each workspace member connects with OAuth, and the AI can read and update what that person can access. It works on all Notion plans; Business and Enterprise add MCP tokens, connected-source search and higher limits. The old open source notion-mcp-server is no longer actively maintained.

  • Where do you find MCP servers, and which can you trust?

    Find MCP servers in the official MCP Registry (in preview), on the vendor's own site or GitHub, and in community lists such as awesome-mcp-servers. None of these vouch for safety: the registry checks who published a server, not what its code does. Prefer the vendor's official server, check the owner, transport, sign-in and write tools, and test it before the team uses it.

  • How do you add MCP servers to Claude Code and share them with your team?

    You add an MCP server to Claude Code with claude mcp add. With --scope project it lands in a .mcp.json file in the repository, so everyone who clones the repo gets the same servers after approving them. Secrets stay out of the file through ${VAR} expansion or per-user OAuth. Admins can fix or filter the servers for the whole organization with managed-mcp.json, managedMcpServers and allow- and denylists.

  • What is the Context7 MCP server, and what does it not do for a team?

    Context7, from Upstash, is an MCP server that gives coding agents current, version-specific documentation and code examples for public libraries, so they rely less on outdated training data. It has two tools, resolve-library-id and query-docs, and runs remotely at https://mcp.context7.com/mcp. As of November 2026 the Free plan includes 1,000 API calls a month and Pro costs $10 per seat a month, with private repositories.

  • How do you add MCP servers to Codex, and can Codex still run as an MCP server?

    Codex reads MCP servers from config.toml: ~/.codex/config.toml for you, or .codex/config.toml in a trusted project for the team. You add one with codex mcp add or a [mcp_servers.<name>] table, and sign in with codex mcp login. The ChatGPT desktop app, Codex CLI and IDE extension share this configuration. Running Codex itself as an MCP server is no longer possible: OpenAI removed the codex mcp-server command.

  • How do you set up MCP servers in Cursor for a whole team?

    Cursor reads MCP servers from two files: .cursor/mcp.json in the project, which you can commit so the team shares it, and ~/.cursor/mcp.json in your home folder for yourself. It supports stdio, SSE and Streamable HTTP, OAuth for remote servers, and ${env:NAME} for secrets. Team admins can distribute shared servers; Enterprise admins can configure an MCP allowlist.

  • Is there an official Gmail and Google Workspace MCP server?

    Yes. Since May 1, 2026 Google offers its own remote MCP servers for Gmail, Drive, Docs, Sheets, Slides, Calendar, Chat, People and Classroom, as a developer preview. Each has its own endpoint, such as https://gmailmcp.googleapis.com/mcp/v1, and uses OAuth 2.0 per user. You need a Google Cloud project and your own OAuth client. Claude and Google Antigravity are documented clients.

  • Local vs remote MCP server: which should a team use?

    A local MCP server is started by your AI app on your own machine and talks over stdio; it typically serves one client and gets its credentials from your environment. A remote MCP server runs on its own over Streamable HTTP, usually with OAuth sign-in, and can serve many people and many AI tools. For a team, remote is the only kind everyone can reach in the same way.

  • How do you use n8n as an MCP server for Claude, ChatGPT or Cursor?

    n8n works with MCP in three ways. The instance-level MCP server, enabled under Settings > Instance-level MCP, lets Claude, ChatGPT, Cursor, Codex and other clients search and run the workflows you mark as available. The MCP Server Trigger node turns a single workflow into its own MCP server. And the MCP Client Tool node lets n8n's own AI agents call outside MCP servers.

  • What is an MCP memory server and how does it work?

    An MCP memory server is a small program that keeps facts outside the AI model and exposes them as tools over the Model Context Protocol. The AI calls those tools to save, search and delete memories, so what it learned survives a new chat. Where that memory lives, and who else can read it, depends entirely on the server you pick.

  • Does the Mem0 MCP server give a whole team shared AI memory?

    The Mem0 MCP server is hosted by Mem0 at mcp.mem0.ai and keeps memories in your Mem0 Platform account, so any MCP client you connect with that account (Claude, Cursor and others) can read and write them. That makes it portable memory across tools. It does not, by itself, turn into team memory with per-person access, learned team rules or your existing documents.

  • How do I set up the official MCP memory server (server-memory) in Claude Desktop?

    Add the server to Claude Desktop's mcpServers config with npx -y @modelcontextprotocol/server-memory (Node.js 24 or later) or the mcp/memory Docker image, then restart Claude. Set MEMORY_FILE_PATH to an absolute path you own, such as ~/memory.jsonl, because the default file sits inside the package folder and can disappear on an update.

  • What is the Graphiti MCP server and is it safe to use for a team?

    The Graphiti MCP server is an experimental MCP server from Zep that gives AI tools access to Graphiti, a knowledge graph that tracks when facts were true. It runs on your own machine with Docker or Python plus a graph database and an LLM API key. Its docs warn it has no authentication, authorization or prompt injection protection, and should not be used as a production deployment.

  • How do I connect an MCP server to Claude, ChatGPT, GitHub Copilot and Grok?

    In Claude and ChatGPT you add a remote MCP server as a custom connector with its URL and sign-in; in Team, Business and Enterprise workspaces an admin or owner does it. In GitHub Copilot you add it in VS Code, and on Business and Enterprise an admin must first allow MCP. Grok supports custom MCP connectors in Grok Business and Enterprise, and remote MCP through the xAI API.

  • Is MCP safe to use in a company, and what controls do we need?

    MCP is a protocol, so it is as safe as the servers you connect and the rights you give them. The official security guidance names concrete risks: token passthrough, confused deputy, SSRF, session hijacking and local servers running malicious commands. A company needs trusted servers only, per-person sign-in with minimal scopes, read-only tools first, human approval for writes, and admin control over which servers each AI tool may use.

All topics →

The first half hour · free

What does your AI forget?

  1. 01You tell us which AI you use and where it keeps starting over.
  2. 02We show what a shared memory changes there, on your own examples.
  3. 03You keep the plan, with or without Bombos.
Piet Baudoin
Piet BaudoinYou talk to Piet, from the Bombos teamDutch team from Utrecht and Nijmegen · data in Europe
We reply the same working day.