"Guide to Developing and Deploying a Web SaaS Application" matching MCP servers:

Enables MCP-capable agents to perform local web research under a declared plan with session budgets, user-approved limit extensions, and guardrails for safe searching, page fetching, content extraction, and audit logging.

Enables AI agents to drive a logged-in ChatGPT web session over CDP, including managing projects and chats, sending messages with file attachments, and retrieving long Pro replies.

Enables AI coding agents to request a second opinion from a different frontier model via a single tool call, with PII scrubbed before transit, live web search, and support for Claude, ChatGPT, Cursor, and Gemini.

Enables AI agents to request an independent second opinion from a different frontier model, with PII scrubbing before transit, live web search grounding, caching, and multi-provider failover.

Orchestrates AI agents through structured markdown documents, enabling multi-agent workflows with automatic context injection and workflow management.

Enables Claude to fetch web pages from the user's own local network connection instead of AI datacenter IPs, including optional authenticated sessions, with verdicts for bot challenges, login walls, and other failures.

Enables Codex to send text tasks to ChatGPT web, await responses, continue conversations, choose reasoning effort, and manage projects through MCP tools on macOS.

Enables building and evaluating LangGraph agents over a protocol-level MCP server to measure failure rates across four agent topologies.

Enables ChatGPT web to use local tools like file reading, command execution, and patch application through an MCP server over OpenAI Secure MCP Tunnel.

An MCP bridge that exposes DeepSeek Harness (DSH) web UI's cordis RPC API as stdio tools, enabling any MCP client to manage workspaces, create/resume sessions, send messages, and fetch session stats.

An unofficial MCP server that lets MCP clients like Codex operate the ChatGPT web interface through a persistent browser, enabling prompt writing, file uploads, message sending, and reading complete responses without using the ChatGPT API.

Enables web search for the pi coding agent using z.ai GLM native search with a DuckDuckGo fallback.

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.

Enables AI agents to control a real browser via MCP tools for web navigation, interaction, page inspection, tabs, forms, files, sessions, and browser automation.

Enables MCP-capable AI harnesses to interact with DeepSeek Web using an authenticated browser session, supporting prompts, deep think, search, image recognition, file attachments, and local SKILL.md injection.

Acts as a bridge between standard REST/OpenAPI web APIs and the Model Context Protocol, dynamically converting REST endpoints into MCP tools for LLM clients like Cursor, Claude Desktop, and n8n.

Pluggable local agent capability hub that gives web-based DeepSeek local Agent capabilities (MCP/file operations/skills/desktop automation) and lets local agents reuse the same tools.

An MCP server for automated interaction with Gemini's web interface using Playwright and LangGraph, enabling task execution, deep research, file uploads, chat management, and dynamic selector verification via MCP tools.

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

Enables MCP clients that lack built-in search — such as Claude Code through Bedrock or local LLM clients — to run web queries against Amazon's operated search index via a Gateway deployed in your own AWS account, with queries never leaving AWS and citation URLs preserved. Supports IAM/SigV4, Cognito, and API key authentication, per-gateway rate limits, IP and domain restrictions, and result filtering by domain and publication date.