"Top trending MCP servers currently popular" matching MCP servers:

Enables AI agents to browse, price, rent, manage, and stop GPUs, top up balances via x402, and register their own accounts, with read-only defaults until spending is explicitly enabled.

An MCP proxy that embeds tool definitions and returns only the top-k semantically relevant tools per query, reducing context usage and routing calls to upstream MCP servers.

A Meta-MCP Server that acts as a tool discovery service, helping AI assistants find appropriate MCP servers from a database of 800+ servers when they need capabilities that aren't currently available.

Search and discover AI agents, skills, prompts, bundles and MCP connectors from the Spark catalog. Seven tools: search assets, get asset details and raw content, list popular assets and categories, check API key usage, and report the outcome of applying fetched content. No API key needed.

MCP server for searching and managing AI agent skills from skills.sh, including search, popular skills, details, and install commands.

Aggregates multiple MCP servers into a single gateway with unified top-level tools, reducing LLM context usage and enabling IDE compatibility by consolidating many tools into fewer interface functions.

MCP server for discovering and installing AI agent skills from agentskill.sh. Search skills across platforms, browse trending skills, and install them with built-in security scanning.

Human-in-the-loop approval gateway for agent tool calls: agents request, policies decide, humans approve via Slack/Discord/web — with an OWASP-LLM-Top-10-tagged audit trail. Self-hostable.

A secure MCP server enabling tool calls (kb_search, read_doc, publish_report) through a zero-trust CapabilityBroker with OWASP LLM Top-10 guardrails and human-in-the-loop approval.

Worker that turns any URL into agent-callable MCP tools, plus a Chrome extension, open-source adapters, and SDKs. On top of that sits an independent A–F trust graph for the whole MCP ecosystem

Enables AI agents to send messages to human users via chat platforms (currently Discord) and optionally wait for human responses, facilitating human-in-the-loop interactions.

A gateway that manages multiple MCP servers behind one endpoint, providing namespaced tool routing and supervision. Currently bundles an Arxiv MCP server for paper search and retrieval.

SkillMesh is a retrieval router for MCP/LLM tool catalogs. It selects top-K relevant expert cards instead of loading every tool, reducing prompt size, by improving tool selection, and lowering token cost. Supports Claude MCP, Codex skills, and OpenAI-style function schemas.

A Model Context Protocol server that allows LLMs to interact with web content through standardized tools, currently supporting web scraping functionality.

A progressive-disclosure gateway for MCP servers that keeps tool lists small by exposing one top-level tool per server, allowing agents to search, list, inspect, and call underlying tools within a selected domain.

A gateway MCP server that exposes selected marc-andre-app functions to Claude in risk tiers, featuring dual authentication, caching, and a kill switch. It requires persistent hosting and is currently stubbed until the backend routes are implemented.

Enables ChatGPT to call an orchestrator agent via a single MCP tool, currently a stub but designed to be replaced with a real agent for task execution.

Enables any MCP client to obtain a typed API for whatever page the user is currently on, then execute the call in their real, logged-in browser session without screenshots, DOM dumps, or accessibility trees entering the caller's context. It grafts WebMCP onto sites that never adopted it, synthesizing tools from ordinary forms and ARIA widgets so a page becomes one validated call instead of many probing steps.

Task board for MCP agents that lives on your desktop: four columns for agents to write into, per-task progress, stall detection and an append-only event log — plus an always-on-top Windows widget so you can watch a run without switching windows. Zero third-party dependencies.

A TypeScript MCP server for launching, tracking, and managing external coding-agent runs across local and remote backends like Codex and Claude Code. It allows top-level agents to orchestrate subagents through tools for spawning tasks, polling events, and handling interactive sessions.

Enables an agent to reconfigure its own harness fleet instead of relying on a human: activating skills just in time, switching MCP servers on and off, and reading or appending master instructions. It also reports fleet status so you can see which harnesses are currently linked.

Acts as a proxy between an MCP client and one or more MCP servers, aggregating upstream tool definitions and exposing only the top-k tools that best match the current task intent. It enables users to shrink large tool catalogs via hybrid retrieval over tool names and descriptions, with intent set through CLI flags or a runtime control tool, while routing tool calls to the correct upstream server.

Drives the local Doubao desktop client through its DevTools interface, letting an agent read and send chats, run task-mode work, hand off a project directory and take it back, and call Doubao-only capabilities like image generation and step traces. Also borrows Doubao's built-in skill set for cross-checking and project handoff, with a UI badge showing which side currently owns the work.

Pools enabled free-tier, keyless, and trial LLM chat routes behind a single OpenAI-compatible endpoint with automatic failover and local usage tracking, so coding agents and OpenAI/Anthropic-compatible tools can run on free models without changing their code. It also exposes provider health, capacity, model inventory, and routing-mode tooling so clients can discover and use whichever chat routes are currently accessible, including some that require no API key.