Read and write local files with configurable access controls.
Editorial picks · stable across releases
Builder favorites — install these first
The MCP servers that have stayed in developers’ daily stacks across multiple catalog refreshes. Stable install-first picks — not what just shipped.
Full GitHub API access: repos, PRs, issues, and code search.
Real-time web search with privacy-focused results.
AI-optimized web search with cleaned, citation-ready results.
Full browser automation: navigate, click, screenshot, and scrape.
Structured step-by-step reasoning for complex problem solving.
Trending now · last 90 days
Trending MCPs
New arrivals and movers — MCP servers gaining traction in the latest catalog refresh. Excludes the long-term Builder Favorites above so the two lists do not mirror each other.
Full GitLab API access: repos, MRs, issues, pipelines, and registries.
Production-grade Postgres MCP with index advice, EXPLAIN, and health checks.
Read and post to Discord channels — community management for AI agents.
Manage Linear issues, projects, and cycles from AI context.
Manage Workers, R2 buckets, DNS, and edge policies on Cloudflare.
Search, inspect, and run Hugging Face models and datasets from an agent.
Browse by use case
Top MCPs by category
Each category is structured for direct answers. Find the right MCP for your workflow.
Developer Tools
MCPs that support coding workflows, IDE and CLI usage, and general developer productivity.
Filesystem & Storage
File access, file management, storage interaction, and content retrieval.
Git & Repo Workflows
Repositories, commits, branches, pull requests, and code-history workflows.
Databases
Querying, reading, writing, and inspecting structured data stores.
Browser Automation
Interactive browser control, navigation, clicking, forms, and scripted browsing.
Web Search
Search-oriented access to public web information.
Web Scraping
Extraction and parsing of content from webpages.
Communication
Email, chat, messaging, and collaboration streams.
CRM
Customer relationship management — contacts, deals, pipelines, and account history.
Project Management
Tickets, sprints, roadmaps, and team task tracking.
Docs & Notes
Knowledge bases, wikis, personal notes, and shared documents.
Inbox triage, drafting, search, and reply automation across email providers.
Meetings & Notes
Meeting recording, transcription, summarisation, and action-item extraction.
Design & Creative
Visual design, presentations, video, audio, and creative-asset generation.
Marketing & SEO
SEO research, keyword tools, email marketing, social listening, and growth platforms.
Sales & Outreach
Prospecting, lead intelligence, contact enrichment, and outbound sales workflows.
Education & Learning
Courses, academic research, tutoring, quizzes, and learning-platform integrations.
Productivity
Personal and team workflow acceleration tools.
Documents & Content
Docs, PDFs, slides, sheets, and content creation and manipulation.
Automation & Workflows
Workflow chaining, triggers, integrations, and process automation.
Analytics & Data
Reporting, dashboards, metrics, analytics, and business data tools.
AI & Machine Learning
Model operations, evaluation, inference, and training-related workflows.
Agent Orchestration
Agent routing, multi-step execution, tool coordination, and orchestration.
Security
Authentication, authorization, secrets, scanning, and security-related access.
Cloud & Infrastructure
Cloud resources, infrastructure operations, deployment surfaces, and infra tooling.
Commerce & Payments
Commerce operations, stores, checkout, purchasing, and payment flows.
Scheduling & Calendar
Calendar operations, appointment logic, and scheduling.
Maps, Location & Weather
Mapping, geo-context, routing, weather, and location-aware services.
Domains
Domain registration, transfer, DNS records, and registrar account management.
Web Hosting
Application hosting, deployments, builds, previews, rollbacks, and environment management.
Browse by industry
MCPs by vertical
Contextual picks for specific domains — where the MCP is used, not just what it does.
Browse by client
Using Claude, Cursor, VS Code, Windsurf, ChatGPT or more?
Jump to a client-specific setup guide — config paths, example mcp.json, and curated top picks.
Claude Desktop
The reference MCP client from Anthropic.
Claude Code
Anthropic's CLI — MCPs wired up from the terminal.
Cursor
Add MCP servers to Cursor in under a minute.
VS Code
Configure MCP servers in VS Code via settings.json.
Windsurf
Windsurf Cascade picks up MCP servers from mcp_config.json.
Zed
Zed exposes MCP servers as context servers in settings.json.
ChatGPT
Use MCP in ChatGPT through the Apps directory or custom remote connectors — verified for 2026.
Codex CLI
OpenAI's coding CLI — MCPs configured via a single TOML file.
Why it matters
Why builders use MCPs
Direct tool access
No copy-pasting. MCPs give models real tool access — read files, run queries, call APIs — without manual context passing.
Faster experimentation
Set up once, use anywhere. Add an MCP to your client and it is immediately available across all conversations.
Better agent workflows
Autonomous agents need real tools. MCPs are how you give an agent the ability to act, not just generate text.
Less manual context
Stop feeding the model file contents manually. MCPs handle retrieval so you can focus on the task.
Quick comparison
Most popular MCPs compared
A high-level comparison of the most widely used MCPs across key categories.
| # | MCP | Tags | Setup | Complexity | Labels | |
|---|---|---|---|---|---|---|
| 1 | Filesystem Read and write local files with configurable access controls. | filesystem, files | 2 min | Low | Official | |
| 2 | GitHub Full GitHub API access: repos, PRs, issues, and code search. | git, github | 5 min | Low | Official | |
| 3 | Context7 Up-to-date library docs pulled directly into your AI context. | coding, documentation | 3 min | Low | ||
| 4 | Memory Persistent knowledge graph memory across AI conversations. | memory, persistence | 2 min | Low | ||
| 5 | Brave Search Real-time web search with privacy-focused results. | search, web | 5 min | Low | ||
| 6 | Postgres MCP Pro Production-grade Postgres MCP with index advice, EXPLAIN, and health checks. | database, sql | 5 min | Medium |
For detailed comparison within a category, see the category pages.
Long-form guides
Buyer's guides — pick the right MCP for the job
Decision trees, compare tables, and gotchas for the queries developers actually type. One guide per high-intent category.
MCP directories compared
Top MCPs vs mcp.so, PulseMCP, mcp.directory, Glama — which directory wins on curation, depth, and citation accuracy.
Best MCPs for databases
Postgres MCP Pro vs Supabase vs SQLite vs warehouse MCPs — read-only contracts, security advisories, and which one to install first.
Best MCPs for research
Tavily, Brave Search, Context7 — when to pick each, how they fail differently, and how to combine them in one agent.
Best MCPs for design-to-code
Figma + design-system context. How to get accurate spacing, tokens, and components from an agent instead of hallucinated CSS.
FAQ
Frequently asked questions about MCPs
Direct answers to common questions about Model Context Protocol tools.
What is an MCP?
MCP (Model Context Protocol) is an open standard that lets AI models connect directly to tools, APIs, and data sources. Instead of copy-pasting context manually, an MCP-enabled model can read files, run queries, search the web, or call APIs on its own.
Why use MCPs?
MCPs eliminate the manual work of feeding context to AI models. They give models direct tool access — filesystem, databases, search, code execution — making AI assistants genuinely useful in developer workflows instead of just chat interfaces.
Which MCPs are best for developers?
For most developers, start with: Filesystem (local file access), GitHub (repo management), Context7 (accurate library docs), and either Postgres or SQLite (database access). These cover the most common use cases with minimal setup.
How do I choose an MCP?
Match the MCP to the task: need the AI to read/write files? Use Filesystem. Need current docs? Use Context7. Need GitHub operations? Use the GitHub MCP. Filter by complexity and setup time to find options that fit your workflow.
Are MCPs useful for AI agents?
Yes. MCPs are core infrastructure for autonomous agents. They give agents the tools to act — executing code, reading files, storing memory, searching the web — rather than just generating text. The Memory and Sequential Thinking MCPs are particularly agent-focused.
Do MCPs work with Claude, Cursor, and other tools?
Yes. MCPs work with any MCP-compatible client. Currently supported clients include Claude Desktop, Cursor, Zed, VS Code (with MCP extensions), and custom agent frameworks using the MCP SDK.
Ready to find your next MCP?
Browse by use case to find the right MCP for your workflow. Every page is structured for fast evaluation.

