"How to allow an LLM to ask questions to a user" matching MCP servers:

An autonomous MCP server that uses the Claude Code CLI to solve coding problems without permission prompts.

Converts AI Skills (following Claude Skills format) into MCP server resources, enabling LLM applications to discover, access, and utilize self-contained skill directories through the Model Context Protocol. Provides tools to list available skills, retrieve skill details and content, and read supporting files with security protections.

A Model Context Protocol server that converts Markdown content to HTML format.

An MCP server that turns a one-sentence idea into working, tested code using a pipeline of six AI agents, with optional deployment.

One-pass agentic inbox triage as an MCP server: fetch unread Gmail → classify (action_needed/fyi/newsletter/noise) → summarize → extract tasks → draft replies as Gmail DRAFTS (never sends) → flag calendar → write a triage report. Four stdio tools (fetch_emails, save_gmail_draft, append_tasks, write_report); the host is the LLM, so it runs keyless in Claude Code. Gmail scopes: readonly + compose

Enables MCP clients to interact with A2A agents through four ordinary tools, translating agent discovery, messaging, task reading, and cancellation between the A2A and MCP protocols, with support for the tasks extension.

MCP server that lets you drive Claude Code sessions hands-free by voice from any MCP client, enabling remote models to send prompts, monitor progress, and get results read back.

An MCP server integration that enables Cursor AI to communicate with Figma, allowing users to read designs and modify them programmatically through natural language commands.

MCP server for qualifying and responding to inbound leads in seconds using a multi-agent AI pipeline.

Enables Codex to clarify requirements via native MCP elicitation controls, supporting single/multiple-choice and free-text questions with recommended answers and a discuss-first option.

Enables users to control the cursor in Figma through verbal commands using an agentic AI agent, streamlining the design process with a new interaction method.

Enables two LLMs to play Tic-Tac-Toe against each other autonomously using a shared tool and an SSE relay. The server facilitates agent-to-agent communication by holding tool responses until the opponent makes a move, managing the game state in real-time.

Enables Claude to use Google Gemini as a secondary AI through MCP for large-scale codebase analysis and complex reasoning tasks. Supports both Gemini Flash and Pro models with specialized functions for general queries and comprehensive code analysis.

Enables Claude Desktop and other AI tools to interact directly with Figma, allowing for powerful AI-assisted design capabilities through natural language commands.

Enables Claude Code agents to perform structured decision-making through TypeSafe Jev, supporting yes/no checks, route selection, scoring, and batched multi-question judgments with confidence values.

Enables agents to break high-level business objectives into dependency-ordered sub-task graphs and then directly generate the finished deliverables, such as financial spreadsheets, executive presentations, and analytical reports. It runs as a zero-dependency Python MCP server over JSON-RPC stdio, adding deterministic authorization, role-boundary validation, and telemetry.

Enables enterprise work agents to decompose high-level business objectives into topological sub-tasks and directly synthesize final deliverables such as financial spreadsheets, executive presentations, and analytical reports.

Enables agents to ask users questions through rich, browser-rendered HTML forms and receive structured answers, supporting approvals, choices, free-text, and form inputs.

Enables any MCP-compatible AI agent to ask humans questions in real time via a mobile-friendly PWA with push notifications, operating local-first without third-party services.

Enables users to request a second opinion from a locally authenticated Claude Code model via MCP tools, supporting asynchronous jobs with restricted tools for safety.