"A task involving critical or creative thinking" matching MCP servers:

Enables MCP clients to batch repetitive web admin work by harvesting queue items read-only, collecting verdicts, and applying them only after human approval, with dry runs, staged rollout, rate limits, a circuit breaker and an audit log. Also provides general browser control — navigating, reading, filling forms and clicking on allowed sites through an already logged-in local browser, pausing for confirmation before irreversible actions.

Enables AI agents to run durable, isolated, and observable browser automation tasks via Playwright, with persistent profiles, recovery, evidence capture, and unattended execution for long-running web workflows.

Enables AI agents to control a Playwright-driven browser through natural-language instructions, including page interaction and live browser task execution.

Enables local web crawling and scraping through MCP, providing tools to fetch pages as Markdown, extract structured data with CSS selectors, and capture full-page screenshots while managing a shared browser instance.

Enables AI assistants to search for vehicles, manage reservations, and access Enterprise Plus loyalty account information through headless browser automation on enterprise.com. It supports the complete car rental workflow including vehicle searching, booking, modifications, and points balance checks.

Read-only MCP server for Tehran Stock Exchange (TSETMC) live market data and authenticated TSE portal pages, enabling quotes, order books, market watch, and session-safe page reads via natural language.

Provides a local MCP server for AI-powered web crawling on Windows, enabling agents to directly crawl and extract web content without requiring WSL.

Enables agents to record a browser task once and replay it across Excel or CSV rows without a model, verifying each item and journaling results so runs can resume and uncertain submissions are held for review.

A powerful Model Context Protocol (MCP) server for scraping Behance.net. Extract projects, user profiles, images, and job listings from Behance's creative community without any API keys or subscriptions.

Connects MCP clients to a user's local Chrome profile via a loopback-only WebSocket, adding isolated agent windows, named task groups, workspace cleanup, and explicit human review prompts.

Enables MCP clients to automate an authenticated browser session for web chat, model and thinking-depth switching, new-chat resets, and local file upload-and-ask workflows.

MCP-native release confidence control plane that turns browser execution into auditable go/no-go decisions by combining business-critical journey context, evidence-heavy QA runs, and risk governance.

Enables MCP agents to drive an already signed-in Chrome or Edge browser through a local extension bridge, controlling existing tabs and opening task-related HTTP(S) tabs without remote debugging or profile copying.

Enables agents to post bounty-backed tasks, browse and inspect a public task board, and pay to execute jobs such as page fetching, link checking, change detection, and claim verification. Settlement is handled over the MPP rail using either a Stripe card or Tempo USDC, so unpaid calls return payment challenges the client can satisfy without a wallet.

The webslinger-mcp server enables users to execute scripts to perform web-based tasks via the webSlinger browser extension. webSlinger is an automation tool where a user demonstrates a task to perform and records a session map which is used to create a script. Such scripts can be run demand, on a schedule, or via the mcp-server.

Browser agents get stuck when a task needs a person. easi.run provides isolated Chrome on your own Linux server through MCP or REST: watch live, take over for login or approval, then hand back. Saved profiles retain browser state. Requires Linux/Docker. AGPL-3.0; server and model costs separate. Code/setup: https://github.com/julianosig/easi.run

You can pay 1 USDC and launch a .xyz subdomain, or 10 USDC and launch a fully hosted .xyz website.

Provides Hermes with an isolated Chromium browser task space enabling structured navigation, semantic snapshots, form interaction, screenshots, uploads, and downloads.

An open-source MCP server that lets AI agents investigate existing websites or applications with a real browser and static analysis, then rebuild functional alternatives through a dependency-ordered, test-driven loop. It provides persistent project state, task-scoped build context, runner execution, and browser-based validation that scores the rebuild's fidelity against the original.

This MCP server lets users capture a website task once, review and edit the generated JSON tool spec, and expose enabled tools to Claude or other MCP clients over stdio or loopback HTTP. It also provides live browser tools for opening, snapshotting, clicking, typing, waiting, and screenshotting pages within safety limits.

Enables agents to discover browser-task skills recorded once by a person and replay them with their own inputs, optionally through a capped no-install public try or a locally-run executor that pauses risky submit, buy, send or delete steps for human approval. Each run yields a signed, hash-chained record whose content hash and Ed25519 signature any client can verify.

MCP server for automating Weavy.ai creative workflows. It provides tools to manage projects, add nodes, configure parameters, run generative nodes, and publish workflows via browser automation.

Enables AI agents to drive your own signed-in ChatGPT web account — asking questions, continuing chats, attaching files, toggling thinking or web search, creating images, transcribing recordings of any length, reading replies aloud, and managing conversations, memories and projects. It runs against a local Chrome tab on chatgpt.com and reports live progress and step-named errors back to the client.

Enables AI agents to drive a real Chrome browser using text-only accessibility-tree snapshots, with tools for navigation, interaction, and persistent task-space management.

Exposes TaskFleet's pay-per-call AI task APIs as MCP tools, including web search, scraping, extraction, translation, code review, summarization, and wallet intelligence. Payments are handled via x402 (USDC on Base) with no accounts or API keys.

Lets an AI assistant read a browser task the user recorded with the ShowOnce Chrome extension (action log and screenshots) and hand back a plan that ShowOnce turns into an n8n workflow.

This task requires analyzing a README, but I'm unable to browse external URLs. Based on the provided README text, the MCP server name is "Agent Search MCP."

Enables MCP clients to control a local Chrome browser, including tabs, navigation, reading, clicking, typing, screenshots, and exclusive per-task control with user-stop capabilities.

Enables ChatGPT to work on a local machine through an MCP connection by inspecting projects, editing files, running bounded terminal and Git commands, driving Chrome or Edge, and reading or writing an already-connected Excel workbook. It invokes the Codex model explicitly only when a task genuinely needs it, so ordinary local tool actions consume no Codex model usage.

Enables controlling a real Chromium browser via MCP tools using structural text snapshots instead of screenshots, supporting both step-by-step interaction and fully autonomous task execution.