DEV Community: curatedmcp

MCP Ecosystem Week 20: Database Tooling Enters the Ring as Giants Dominate Discovery

curatedmcp — Mon, 11 May 2026 10:17:57 +0000

Originally published at curatedmcp.com/blog/week-2026-20

MCP Ecosystem Week 20: Database Tooling Enters the Ring as Giants Dominate Discovery

The MCP ecosystem continues its steady expansion with the addition of Local-YDB, an unofficial server that brings local YDB database management into the MCP fold. Meanwhile, the weekly trending list reveals a clear pattern: integration servers backed by major platforms—OpenAI, GitHub, Figma, and Anthropic—continue to capture developer attention at scale.

This Week in MCP

We added one new server this week:

Local-YDB unofficial mcp server (FREE): A community-built server for managing local YDB database instances. YDB, Yandex's distributed SQL database, joins the growing roster of database-specific MCP tools as developers seek to integrate data layer operations directly into their AI workflows.

The addition pushes the verified catalog to 70 servers, with 69 remaining free. No new playbooks were added this week, keeping the focus on server infrastructure expansion rather than orchestration patterns.

Trending Servers

GitHub Copilot MCP continues its dominant run, commanding 98K views this week. The server channels GitHub's code intelligence—completions, explanations, and review capabilities—into any MCP client, making it an obvious draw for developers working in Cursor and other AI IDEs. It's the bridge between GitHub's ecosystem and broader AI agent workflows.

OpenAI MCP lands at 87K views, offering official access to GPT-4o, DALL-E, Whisper, and Embeddings APIs through MCP. The steady demand signals that developers want to orchestrate multiple OpenAI capabilities within unified agent contexts rather than making isolated API calls.

Figma MCP rounds out the top three with 82K views, serving design engineers and teams building AI-assisted workflows around design systems. Direct access to design files, components, and design tokens from within an agent drastically reduces context-switching overhead—a compelling value proposition that the numbers clearly reflect.

The pattern here is worth noting: first-party integrations from well-known platforms capture the lion's share of attention. GitHub, OpenAI, Figma, and Anthropic collectively pull roughly 420K views across the top five slots. Developer friction and discoverability favor the names they already trust.

Quick Take

The 100% free-tier catalog (69 of 70 servers) masks an important reality: pricing structures around MCP are still forming. No commercial/premium servers have gained traction yet, suggesting the ecosystem remains in the "free tooling to build the foundation" phase. What's notable is how developers are using free servers—they're threading together specialized capabilities (database tools, design platforms, code hosts) into composite agents rather than adopting monolithic platforms. This suggests MCP's real value isn't in individual integrations, but in the ability to mix and match. The Local-YDB addition is a quiet marker: database teams, not just SaaS platforms, are beginning to see MCP as infrastructure worth supporting. Watch for more domain-specific tooling (observability, data, infrastructure) to follow as the pattern becomes clearer.

Browse all 70 verified servers at CuratedMCP.

Private MCP registry your CISO has been asking for

curatedmcp — Sun, 10 May 2026 20:32:13 +0000

npm private registry, but for MCP. Give every engineer one governed source of truth for the AI agent tools they use in Claude, Cursor, Windsurf, and Copilot — with RBAC, audit logs, and procurement-friendly billing.

check it up:
https://www.curatedmcp.com/enterprise/private-registry

PagerDuty MCP: Let AI Handle Your On-Call Chaos

curatedmcp — Sun, 10 May 2026 10:45:34 +0000

Install guide and config at curatedmcp.com

PagerDuty MCP: Let AI Handle Your On-Call Chaos

Operations teams spend too much time context-switching between incident dashboards, chat, and email. PagerDuty MCP is the official integration that gives Claude, Cursor, and other AI agents direct access to your incident response workflows—so they can create incidents, manage alerts, update on-call schedules, and keep responders aligned without human hand-offs.

What It Does

PagerDuty MCP bridges the gap between your AI agent and your incident management system. Instead of manually toggling between tools, you can now ask Claude to create an incident, check who's on-call, escalate a critical alert, or annotate a timeline—and the agent executes those commands directly in PagerDuty.

The server unlocks core capabilities: create and update incidents with custom urgency and priority levels, list active on-call users and rotation schedules, manage escalation policies and team assignments, query historical alerts and response timelines, annotate incidents with context or responder notes, and trigger incident actions programmatically from plain-language AI commands.

This is especially powerful for automation. When your monitoring system fires a critical alert, an AI agent can automatically open a PagerDuty incident, notify the right on-call engineer, and add diagnostic context—all in seconds. Or during an active incident, your agent can pull recent alert history, suggest escalation paths, and track response metadata without anyone leaving their terminal.

How to Install

npx -y @pagerduty/mcp

Add the server to your Claude Desktop config:

{
  "mcpServers": {
    "pagerduty-mcp": {
      "command": "npx -y @pagerduty/mcp",
      "env": {
        "PAGERDUTY_API_TOKEN": "your-api-token-here"
      }
    }
  }
}

You'll need a valid PagerDuty API token with appropriate scopes. Get yours from your PagerDuty account settings.

Real-World Use Cases

Incident triage at scale: Ask Claude to review your last 50 incidents, identify patterns (e.g., recurring database timeouts), and propose new escalation policies without manually digging through the UI.
Automated alert response: When a monitoring tool detects anomalies, an AI agent creates a PagerDuty incident, pulls the on-call roster, checks recent similar incidents, and sends a pre-populated summary to the responder.
On-call coordination: Query who's currently on-call across multiple services, check schedule conflicts, and let Claude suggest load-balancing changes based on response time trends.

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.

Splunk MCP: Query your observability stack directly from Claude

curatedmcp — Sat, 09 May 2026 10:45:37 +0000

Install guide and config at curatedmcp.com

Splunk MCP: Query your observability stack directly from Claude

Splunk MCP is the official Model Context Protocol server that bridges AI agents to your Splunk instance. Instead of context-switching between Claude and your monitoring platform, you can now ask your AI assistant to search logs, inspect alerts, correlate incidents, and surface root causes—all without leaving your chat.

What It Does

Splunk MCP gives Claude and other AI agents direct access to your observability and security data. You can run SPL (Splunk Processing Language) queries, retrieve event results, inspect logs across services, and access alert timelines. The server also lets you search saved dashboards, manage data inputs, and correlate security events with operational metrics.

The key unlock: your AI agent can now triage incidents autonomously. Feed it vague symptoms ("why did latency spike at 3pm?"), and it can query your Splunk instance, pull relevant logs and metrics, cross-reference alert history, and surface likely root causes. For SRE and SecOps teams drowning in data, this cuts investigation time significantly.

How to Install

Install via npm:

npx -y @splunk/mcp

Then add Splunk MCP to your Claude Desktop config:

{
  "mcpServers": {
    "splunk-mcp": {
      "command": "npx -y @splunk/mcp",
      "env": {
        "SPLUNK_URL": "https://your-splunk-instance.com",
        "SPLUNK_USERNAME": "your_username",
        "SPLUNK_PASSWORD": "your_api_token"
      }
    }
  }
}

Replace placeholders with your Splunk credentials and instance URL. Full setup guides for Cursor, Windsurf, and other clients available in the official docs.

Real-World Use Cases

Incident triage at speed: "Hey Claude, I got a PagerDuty alert for high error rates. Search Splunk logs from the last 30 minutes and tell me which service is failing." Claude queries your instance, correlates events, and suggests remediation steps.
Root cause analysis: "Why did our database queries slow down yesterday around 2pm?" Claude pulls metrics, compares baselines, and surfaces anomalies you'd normally miss.
Security event correlation: "Find all failed login attempts from the last 48 hours and correlate them with successful logins from unusual IPs." Claude runs complex multi-field searches without you writing SPL.

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.

ServiceNow MCP: Automate ITSM workflows without leaving your AI agent

curatedmcp — Fri, 08 May 2026 10:45:35 +0000

Install guide and config at curatedmcp.com

ServiceNow MCP: Automate ITSM workflows without leaving your AI agent

ServiceNow MCP is the official integration that lets Claude, Cursor, and other AI agents interact directly with ServiceNow's ITSM platform. Instead of context-switching between your IDE and the ServiceNow portal, you can now orchestrate incidents, change requests, and approvals entirely through natural language prompts in your agent.

This is particularly useful for ops teams building automation workflows or developers who need to integrate service desk operations into their deployment pipelines. The server speaks both directions—your AI agent can read incident data, search the knowledge base, and write changes back to ServiceNow in real time.

What It Does

ServiceNow MCP unlocks full ITSM lifecycle management through your AI agent. You can create and update incidents, problems, and change requests; search knowledge base articles to surface solutions during troubleshooting; manage approval workflows and SLAs; query asset and configuration item data; and access service desk analytics to identify trends.

The server acts as a direct API bridge, so your agent can handle repetitive ticket routing, auto-assign based on skill tags, fetch related configuration items during incident triage, or even generate reports from trend data. Think of it as giving Claude or Cursor read-write access to your entire service desk—without the browser tab.

How to Install

Run this single command:

npx -y @servicenow/mcp

Then add the server to your Claude Desktop configuration (or Cursor/Windsurf equivalent):

{
  "mcpServers": {
    "servicenow-mcp": {
      "command": "npx -y @servicenow/mcp",
      "env": {
        "SERVICENOW_INSTANCE": "your-instance.service-now.com",
        "SERVICENOW_API_KEY": "your-api-key"
      }
    }
  }
}

Replace SERVICENOW_INSTANCE and SERVICENOW_API_KEY with your credentials. See the official docs for authentication details.

Real-World Use Cases

Incident triage in Cursor: Paste an error log, ask Claude to search the knowledge base for solutions, and auto-create an incident with relevant config items attached—all without leaving your editor.
Change request automation: Describe a deployment, let your AI draft a change request with SLA calculations, and route it to the approval queue based on impact level.
On-call escalation: Build a Slack bot or CLI tool that queries open incidents, fetches analytics, and drafts summaries for your incident commander—all via natural language.

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.

MCP Ecosystem Week 19: Data Intelligence and Safety Tools Lead New Arrivals

curatedmcp — Mon, 04 May 2026 10:38:54 +0000

Originally published at curatedmcp.com/blog/week-2026-19

MCP Ecosystem Week 19: Data Intelligence and Safety Tools Lead New Arrivals

The MCP marketplace continues to expand with four new servers landing this week, reflecting a clear market demand for data accessibility and content safety. Week 19 brings tools spanning SEO analytics, conversational UI, marketing data aggregation, and threat detection—a diverse mix that signals the ecosystem is moving beyond developer-centric integrations toward enterprise and consumer-facing use cases.

This Week in MCP

Four new verified servers joined the catalog:

DataForSEO MCP Server opens SEO and SERP data to any MCP client. For developers building competitive analysis or content tools, this eliminates the friction of API documentation and token management.

Perspective AI replaces form-based interactions with adaptive AI conversations. The shift from static inputs to natural dialogue is gaining traction—this server makes that pattern pluggable for any agent.

Windsor.ai MCP Server aggregates marketing, sales, and customer data across 325+ sources and exposes it via plain English queries. This is a meaningful consolidation play: instead of agents juggling multiple vendor SDKs, they get a unified interface.

Tuteliq detects grooming, bullying, fraud, and 16+ threat types across text, voice, image, and video. Content moderation and safety are increasingly non-negotiable for production AI systems, and multimodal threat detection filling a real gap.

All four are available free—a continued pattern in the ecosystem.

Trending Servers

The top three trending servers by page views this week reveal what developers are actively integrating:

GitHub Copilot MCP (98K views) leads. The appeal is clear: developers want Copilot's code intelligence—completions, explanations, reviews—available outside the IDE, in their own MCP workflows. It's a portability play.

OpenAI MCP (87K views) maintains second place. Official integrations matter. Having GPT-4o, DALL-E, Whisper, and Embeddings available as MCP resources means developers can route AI workloads through standardized tooling rather than managing OpenAI SDK fragmentation.

Figma MCP (82K views) rounds out the top three. Design-to-code and design system integration are high-friction problems. Direct access to design files, components, and tokens within an AI agent workflow is a legitimate productivity unlock—and the traffic reflects that need is real.

The pattern: developers prioritize integration servers that reduce context switching and vendor lock-in.

Quick Take

The all-free pricing on this week's new arrivals (and 67 of 68 total servers) reflects a strategic choice by vendors: expand adoption fast, own the integration layer, monetize upstream or at the platform level. But there's a secondary story: we're seeing category clustering around data intelligence (SEO, marketing, CRM aggregation) and safety/moderation. These are problems that individual AI developers can't easily solve in-house. The marketplace is becoming a repository of specialized, repeatable capabilities—the opposite of monolithic platforms. That's healthy ecosystem behavior.

Browse all 68 verified servers at CuratedMCP.

Okta MCP: Let Your AI Agent Handle Identity Management

curatedmcp — Sat, 02 May 2026 10:12:28 +0000

Install guide and config at curatedmcp.com

Okta MCP: Let Your AI Agent Handle Identity Management

Identity and access management shouldn't require manual ticket shuffling. Okta MCP is the official Okta integration for AI agents—it lets Claude, Cursor, or Windsurf directly manage users, groups, authentication workflows, and security events within your Okta tenant.

If you're managing identity infrastructure, this means your AI can provision users, enforce access policies, investigate authentication failures, and respond to security incidents without you leaving your IDE or chat window.

What It Does

Okta MCP grants AI agents safe, scoped access to your identity workflows. Your agent can:

Manage users at scale: create, update, deactivate accounts, reset passwords, and manage recovery rules
Control access: assign users to groups and applications, manage roles, and inspect sign-on policies
Monitor security: query authentication events, check MFA status, and analyze access patterns
Automate OAuth/OIDC: manage client configurations and token policies

The server is built for security and IT teams who want AI to reduce toil—faster user onboarding, quicker incident response, and easier policy enforcement. Everything is gated by your Okta API token, so permissions stay locked to what you've explicitly authorized.

How to Install

npx -y @okta/mcp

Then add it to your Claude Desktop config:

{
  "mcpServers": {
    "okta-mcp": {
      "command": "npx -y @okta/mcp",
      "env": {
        "OKTA_DOMAIN": "your-org.okta.com",
        "OKTA_API_TOKEN": "your_api_token_here"
      }
    }
  }
}

Grab your API token from Admin → API → Tokens in your Okta console. Full setup guides for Cursor and Windsurf are in the docs.

Real-World Use Cases

Onboarding sprint: Ask Claude to provision 12 new contractors, assign them to the engineering group, activate their Slack app, and send a summary—all in one prompt.
Incident response: When a security alert fires, ask your agent to pull MFA status for a suspicious user, review their recent sign-ons, and deactivate sessions—no context switching.
Access review automation: Have your agent query who has admin roles, who hasn't logged in this quarter, and which legacy OAuth clients are still active—then generate a compliance report.

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.

I shipped @curatedmcp/launcher — the MCP hub

curatedmcp — Fri, 01 May 2026 21:27:19 +0000

I built mcp launcher because I kept re-pasting the same MCP server configs
into Claude Desktop, then again into Cursor, then again into Windsurf.

@curatedmcp/launcher is one entry you put in any agent's MCP config. It spawns each registered MCP server as a child process over stdio and
aggregates their tools (prefixed with the server's slug). Add a server
once with launcher add <slug> and it works everywhere.

npx @curatedmcp/launcher init
npx @curatedmcp/launcher add github-mcp --env GITHUB_TOKEN=ghp_...

Everything is local — ~/.curatedmcp/stack.json is hand-editable JSON.

No cloud sync, no account, no auth keys. MIT, ~62 KB unpacked, single
dependency (the official MCP SDK).

The catalog of servers comes from a free index at curatedmcp.com that
I've been building. Source: https://github.com/oneprofile-dev/mcp-launcher

What's missing that you'd want? Looking specifically for: failure
isolation feedback, multi-machine sync (or whether anyone wants it),
and whether the slug-prefix on tool names breaks any agent's UI.

Live @ https://www.curatedmcp.com/launcher

Appreciate your thoughts and questions. Thank you, curatedmcp.com

Five Hard Problems in the MCP Ecosystem

curatedmcp — Fri, 01 May 2026 15:30:40 +0000

A field report from CuratedMCP — what we've learned shipping infrastructure for 10,000+ MCP servers, and what the protocol still can't do at production scale.

Reading time: 9 minutes
Audience: Engineers, security leaders, and AI infrastructure builders

The Model Context Protocol turned one year old in November 2025. In that year it went from a niche Anthropic experiment to the integration layer underneath Claude Desktop, Cursor, Windsurf, GitHub Copilot, OpenAI Agents, and Gemini CLI. There are now over 10,000 active MCP servers and 97 million monthly SDK downloads. The protocol won.

What no one writes about is what happened to the people who tried to build serious infrastructure on top of it.

We've spent the last 90 days running CuratedMCP — a curated marketplace, security auditor, and governance layer for MCP servers. We've shipped npm packages that thousands of developers have installed. We've seen what works at protocol level and what falls apart the moment a real enterprise tries to deploy it. This is not a marketing post. This is a field report on five problems the MCP community is going to have to solve in 2026, and what we've built so far to address each one.

If you're a developer wondering why your MCP setup keeps breaking, a CISO trying to figure out how to govern this, or an investor evaluating the space — read on. The protocol is not the moat anymore. What's built on top of it is.

Problem 1: The discovery layer is broken, and it's getting worse

In early 2025 there were a few hundred MCP servers, scattered across GitHub repos. Today there are more than 10,000. The Anthropic-blessed registry alone passed 2,000 entries in three months. Most of these servers are unmaintained forks, abandoned proofs-of-concept, or duplicate implementations of the same integration.

The problem this creates is invisible until you try to use it. A developer wanting to "connect Claude to Postgres" today has to evaluate roughly seven different Postgres MCP servers, three of which have working install instructions, two of which support the latest spec, and exactly one of which has been updated in the last 60 days. The signal-to-noise ratio is collapsing in real time.

Auto-scraping directories made this worse, not better. They optimize for breadth — "we have the most servers" — and the result is a search experience that surfaces junk alongside production-grade tools, with no way for a developer or a security team to tell the difference.

What we've learned: Curation is the only fix. Every server in our catalog passes a manual review — does it install cleanly, does it have working documentation, does its security posture make sense, is it maintained. That review takes about 20 minutes per server. We list around 70 servers, not 10,000. The customers we have asked us to do less, not more — they wanted "the five postgres MCPs that actually work," not "every Postgres MCP on GitHub."

This is the same insight that got Hugging Face to a $4.5B valuation: Hugging Face hosts roughly a million models, but its product-market fit came from the Transformers library — a curated set of 30 pre-trained models with consistent APIs that actually worked. Breadth is not the moat. Trust is.

The harder problem: curation does not scale linearly with team size. We are working on a hybrid model — community-submitted servers with automated verification (install test, dependency scan, behavior analysis) plus a "Verified by CuratedMCP" badge for human-reviewed entries. The badge is the trust signal; the rest is searchable but unranked.

Problem 2: Distribution lives inside the AI client, not on a website

Here is something the MCP marketplace category got fundamentally wrong, including us, for our first month: developers do not browse marketplaces.

We can prove this with our own analytics. Our marketplace pages get a fraction of the traffic our developer-tool downloads do. The MCP Auditor — a free CLI that scans developer machines for vulnerable MCP configurations — has hundreds of organic npm installs. Nobody promoted it. The traffic to its marketplace landing page is a tenth of those installs.

The lesson: developers find tools by searching npm, by typing into Claude, by reading what their tech lead pasted into Slack. They don't open a tab and go marketplace-shopping. The distribution channel for MCP infrastructure isn't the marketplace website — it's the AI client itself.

That's why we shipped the Launcher: an MCP server whose only job is to find and install other MCP servers. You add one line to your Claude config, restart, and then ask Claude in plain English: "Find me an MCP server for Postgres." The launcher returns a curated config snippet. You paste it. Done.

This pattern — meta-MCP, an MCP server that helps you install MCP servers — is the right shape of distribution for this category. It puts our catalog directly inside the workflow of every Claude, Cursor, and Windsurf user, where they actually live. The marketplace website becomes a dashboard for something that lives in their environment, the same way the GitHub website is a dashboard for something that mostly happens in git.

Problem 3: Shadow MCP is the new shadow IT, and security teams are about to find out

Every developer in your organization is now adding MCP servers to their Claude or Cursor config. Most of those servers are pulled from GitHub or npm with zero security review. They run as local processes with full access to whatever credentials the developer has on their machine — production database keys, AWS secrets, Stripe tokens.

Three things make this uniquely difficult to govern:

It's invisible to network monitoring. MCP servers communicate with the AI client over stdio, not HTTP. There is nothing for a firewall or DLP to inspect.

It's invisible to procurement. A developer doesn't fill out a form to install an npm package. There is no contract, no DPA, no terms-of-service review. The "vendor" is an individual on the internet.

It bypasses endpoint posture. EDR sees node.js running and moves on.

We've talked to half a dozen platform engineering leads at mid-sized companies in the last month. None of them know how many MCP servers their engineers have installed. All of them suspect the answer is "more than fifty, probably some are bad." None have a remediation plan. This is the procurement crisis of mid-2026, and it will be the dominant security conversation for the second half of the year.

What we've built: the MCP Auditor — a free, open-source CLI that scans Claude Desktop, Cursor, and Claude Code configurations for credential leaks, filesystem exposure, and unverified servers. It runs locally. It generates a markdown report. It is the smallest possible useful tool in this space, and it's the one that has had real organic traction.

What we're building: a private MCP registry that gives security teams a single approved catalog their developers can install from. RBAC, scoped API keys, audit logs, instant revocation. The same architectural pattern GitHub Enterprise applied to source code, Okta applied to SaaS auth, and Snyk applied to open-source dependencies. The MCP equivalent does not yet exist at scale, and the demand is about to compound.

Problem 4: The protocol has no answer for production-scale identity

This is the technical problem that gets discussed least and matters most. The MCP team's own 2026 roadmap calls out three protocol-level gaps: identity propagation, adaptive tool budgeting, and structured error semantics.

In practice, here's what that means: when a Claude agent on a developer's machine calls an MCP server that calls a downstream API, who is the request actually being made as? Today, the typical answer is "whatever long-lived API key the developer pasted into the config file when they installed the server six months ago." That key has the developer's full permission scope. It is rarely rotated. It often outlives the project.

A 2025 survey found that 53% of community-built MCP servers rely on static API keys with no rotation discipline. That's not a problem in a hobby setup. In a 500-person engineering org, it's the next major credential-leak incident waiting to happen.

The fix the protocol should support — and is finally moving toward — is OAuth-style authorization with scoped, short-lived tokens, resource indicators that limit token reuse, and explicit consent flows when an MCP server is asked to do something privileged. The 2025-06-18 spec revision adds the framework. Most servers still don't implement it. The migration will take 12-18 months across the ecosystem, and during that window, the surface area for credential abuse grows every week.

What we've built: a configuration scanner that flags every static API key in a user's config and recommends OAuth alternatives where they exist. We're also working on a hosted credential broker — short-lived, scoped tokens issued just-in-time when an MCP server needs them, rather than long-lived secrets sitting in plain text on developer machines. This is the kind of infrastructure that exists for SaaS apps (Vault, Doppler) and doesn't exist yet for MCP. It will.

Problem 5: There is no monetization story for the people building MCP servers

The dirty secret of the MCP ecosystem is that almost no one building MCP servers is making money from them.

Let's actually count. There are 10,000+ active MCP servers. Of those, maybe 200 are built by companies (Stripe, Cloudflare, GitHub, Hugging Face, Notion, etc.) where the MCP server is a free distribution channel for their core SaaS product. The other 9,800 are individuals, side projects, and small teams who shipped something useful and have no path to revenue from it.

This is bad for the ecosystem in two ways. First, the people doing the most interesting niche work — a Datadog MCP, a Splunk MCP, a Salesforce MCP — burn out within six months because nobody is paying them. Second, no monetization means no commitment, which means abandoned servers, broken installs, and the discovery problem from the top of this article.

The Apify model — a marketplace that takes a 20% platform fee and gives developers 80% of every subscription — is what should exist for MCP. It is what we have built. But the harder problem is not the platform mechanics; it's that buyers don't yet understand they should pay for an MCP server, and sellers don't yet understand they can charge.

This is a category-creation problem, not a product problem. Hugging Face had to teach the world that researchers should pay $9/month to host private datasets. Stripe had to teach the world that developers should pay 2.9% to accept payments instead of doing it themselves. Someone in the MCP ecosystem has to teach the world that an MCP server which saves an engineering team 10 hours a week is worth $29/month.

We're trying. Our seller program offers an 80/20 split with automatic Stripe Connect payouts. The number of paid subscriptions is small but growing. The first paid subscription happened three weeks ago. The second happened last week. This is the slow part of building infrastructure.

What this means for the next 12 months

The MCP ecosystem is in a strange phase. The protocol has reached escape velocity. The clients have all adopted it. The number of servers is exploding. And almost everything around the protocol — discovery, security, governance, monetization — is still being built by hand by a small number of teams, in real time, in public.

If you are a developer reading this: install the Auditor, scan your config, see what's actually running on your machine. We give it away because the security data we get back makes the catalog better.

If you are a security leader reading this: assume your engineering org has more MCP servers than you think it does, and that none of them have been reviewed. The cheapest first action is the same Auditor scan, run across your developer machines.

If you are a builder reading this: the MCP server you've published on GitHub is reaching almost no one through that channel. List it on a curated marketplace. Try charging for it. Find out whether anyone will pay. The information from that experiment is worth more than the GitHub stars.

If you are an investor or strategic partner reading this: the layer between the AI client and the actual tools is the next great infrastructure category. MCP is the protocol layer. What gets built on top of it is the thing that compounds.

We are CuratedMCP. We're building the system of record for MCP. We're early. Most of the work above is still in progress. But these are the five problems we wake up thinking about every day, and we'd rather write about them honestly than pretend we have everything figured out.

Find us at curatedmcp.com. The Auditor is at npx @curatedmcp/auditor. The Launcher is at npx @curatedmcp/launcher. The enterprise pilot is open. The conversations we're not having yet are the ones we'd most like to have — drop a line at founders@curatedmcp.com if you've hit any of the problems above.

The protocol is built. The infrastructure on top of it is what we're racing to figure out. Same as 1995. Same as every important technology shift since.

About the author: The CuratedMCP team builds the system of record for the Model Context Protocol ecosystem — a curated marketplace of 70+ verified MCP servers, free open-source security tooling (@curatedmcp/auditor), and an enterprise governance layer for organizations standardizing AI agent tools at scale.

Try it:

npx @curatedmcp/auditor — free MCP security scan
npx @curatedmcp/launcher — install MCP servers from inside Claude
curatedmcp.com/marketplace — verified catalog
curatedmcp.com/enterprise/pilot — 30-day enterprise pilot

Published 2026 · MIT-style content license — feel free to repost with attribution.

www.curatedmcp.com

IBM Cloud MCP: Control Your IBM Cloud Infrastructure From Claude

curatedmcp — Fri, 01 May 2026 10:06:17 +0000

Install guide and config at curatedmcp.com

IBM Cloud MCP: Control Your IBM Cloud Infrastructure From Claude

Managing IBM Cloud resources usually means jumping between dashboards, CLIs, and documentation. IBM Cloud MCP changes that by giving AI agents direct access to your infrastructure, databases, and AI services. Built by IBM for enterprise teams, this official server lets you provision compute, query data, and inspect security policies without leaving your Claude chat or code editor.

What It Does

IBM Cloud MCP is a Model Context Protocol server that bridges Claude, Cursor, and other AI agents to IBM Cloud's full platform. Instead of manual API calls or console clicks, you can ask Claude to spin up Kubernetes clusters, deploy serverless functions, or query Watson AI models directly.

The server handles compute provisioning (Kubernetes, VMs, Cloud Functions), data operations (Cloudant, Db2, Databases for PostgreSQL), and enterprise workflows like identity management and security audits. You get real-time access to logs, monitoring alerts, and metrics across regions. It's designed for teams managing hybrid cloud workloads—IBM Cloud alongside on-premises systems.

For developers, this means faster iteration on cloud-native apps, faster infrastructure troubleshooting, and the ability to automate repetitive cloud tasks through natural conversation.

How to Install

Install via npm:

npx -y @ibm/mcp

Then add it to your Claude Desktop config file (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "ibm-cloud-mcp": {
      "command": "npx -y @ibm/mcp",
      "env": {
        "IBM_CLOUD_API_KEY": "your-api-key-here"
      }
    }
  }
}

Grab your IBM Cloud API key from the IAM dashboard, set it as an environment variable, and restart Claude.

Real-World Use Cases

Deploy and scale an app in seconds: "Deploy my Node.js app to IBM Cloud Kubernetes with 3 replicas across Dallas and London regions, then show me the logs."
Debug database issues with context: "Query my Db2 instance for slow queries in the last hour, then suggest indexes based on the execution plans."
Audit security posture: "List all IAM policies for my service accounts, identify overly permissive roles, and generate a compliance report."

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.

Adobe MCP: Connect Claude to Photoshop, Acrobat, and Experience Cloud

curatedmcp — Thu, 30 Apr 2026 10:06:16 +0000

Install guide and config at curatedmcp.com

Adobe MCP: Connect Claude to Photoshop, Acrobat, and Experience Cloud

Adobe just released an official MCP server that plugs your AI agent directly into Creative Cloud and Experience Cloud. If you're building workflows that need to generate assets, manage PDFs, or pull marketing analytics—this cuts out the manual handoff.

What It Does

Adobe MCP gives Claude, Cursor, and other AI agents direct access to Adobe's creative and marketing tools. You can now:

Create and edit images in Photoshop programmatically—useful if you're generating variations, applying filters, or building batch workflows
Generate and manage PDFs via Acrobat—ideal for automated document creation, form filling, or report generation
Query Experience Cloud for campaign and analytics data—pull reporting directly into your AI context
Manage Creative Cloud assets and access libraries, templates, and brand-approved styles
Export finished work with delivery URLs—no manual export steps

This is officially supported by Adobe, so you're integrating with their actual APIs, not scraping or reverse-engineering. The server runs on your machine or server and communicates with your Adobe account.

How to Install

npx -y @adobe/mcp

Then configure it in your Claude Desktop config (or Cursor/Windsurf):

{
  "mcpServers": {
    "adobe-mcp": {
      "command": "npx -y @adobe/mcp"
    }
  }
}

You'll need Adobe API credentials (client ID and secret from the Adobe Developer Console). Full setup docs at developer.adobe.com/mcp.

Real-World Use Cases

Marketing automation: Ask Claude to pull last month's campaign performance from Experience Cloud, then generate a summarized PDF report with embedded charts—all in one prompt.
Batch creative generation: Feed Claude a product catalog and ask it to generate hero images for each item in Photoshop, apply your brand template, and export URLs for your website.
Document workflows: Automate contract or invoice generation by having Claude populate Acrobat templates with data, then export signed PDFs ready for archival.

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.

Oracle Cloud MCP: Automate OCI infrastructure through AI agents

curatedmcp — Wed, 29 Apr 2026 10:06:16 +0000

Install guide and config at curatedmcp.com

Oracle Cloud MCP: Automate OCI infrastructure through AI agents

Oracle Cloud MCP is the official integration that connects AI agents like Claude and Cursor directly to Oracle Cloud Infrastructure (OCI). Instead of switching between terminals and the OCI console, you can now ask your AI assistant to provision instances, manage databases, handle storage, and configure security—all within your IDE or chat interface.

What It Does

This server bridges the gap between AI agents and OCI operations. It gives Claude, Cursor, and Windsurf read and write access to your OCI environment through the Model Context Protocol. Your AI can:

Compute: Spin up, terminate, and monitor compute instances with different shapes and configurations
Storage: Create and manage Object Storage buckets, upload objects, and handle lifecycle policies
Databases: Query Autonomous Database instances, manage DB systems, and inspect schemas
Networking: Inspect Virtual Cloud Networks (VCNs), review security lists, and audit firewall rules
Identity & Access: Create IAM users, attach policies, and manage permissions programmatically
Security & Compliance: Pull audit logs and generate security reports for governance

It's built for enterprise teams who run OCI and want to shift repetitive infrastructure tasks to AI automation.

How to Install

Install via npm:

npx -y @oracle/mcp

Then add it to your Claude Desktop configuration:

{
  "mcpServers": {
    "oracle-cloud-mcp": {
      "command": "npx -y @oracle/mcp",
      "env": {
        "OCI_CONFIG_FILE": "~/.oci/config",
        "OCI_PROFILE": "DEFAULT"
      }
    }
  }
}

Make sure your OCI credentials are configured locally before connecting.

Real-World Use Cases

On-demand infrastructure scaling: Ask Claude to check current compute usage, then provision additional instances if load is high—all without leaving your editor
Database troubleshooting: Query Autonomous Database logs, inspect connection pools, and get optimization suggestions from your AI in real time
Audit and compliance: Generate security reports, review IAM policies, and pull audit logs to validate your infrastructure against compliance requirements

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.