DEV Community: Rege

Connect MCPs to Your AI Copilot/Chatbot.

Rege — Thu, 05 Mar 2026 09:23:59 +0000

TL;DR: MCP (Model Context Protocol) lets AI connect to external tools — most AI apps like Claude, Cursor support it now. We brought it to web-based AI Copilots, with support for interactive UI components.

We added MCP support to Copilot SDK.

What this means: Your AI copilot can now create Linear issues, query Supabase databases, check Sentry errors, manage Jira tickets — directly from the chat. Users ask, the copilot acts.

MCP (Model Context Protocol) is how AI apps connect to external tools. Claude, Cursor, Windsurf — they all support it. Services like Supabase, MongoDB, Sentry, Linear, Atlassian, AWS, Netlify and many more already provide MCP servers.

Now your AI copilots can connect to them too.

Basic Setup

Add MCP servers to your copilot:

import { CopilotProvider } from "@yourgpt/copilot-sdk/react";
import { CopilotChat } from "@yourgpt/copilot-sdk/ui";

<CopilotProvider
  runtimeUrl="/api/chat"
  mcpServers={[
    {
      name: "my-tools",
      transport: "http",
      url: "https://your-mcp-server.com/mcp",
    },
  ]}
>
  <YourApp />
  <CopilotChat />
</CopilotProvider>

The SDK handles connection, tool discovery, and execution.

Quick Start with MCP360

Want to try MCP without setting up individual services? MCP360 gives you 100+ tools through a single connection:

Web search — Google, YouTube, News
Web scraping — extract data from any page
Lead tools — email verification, company lookup
SEO — keyword research, SERP tracking
Maps — places, directions, reviews

Connect with one URL:

<CopilotProvider
  runtimeUrl="/api/chat"
  mcpServers={[
    {
      name: "mcp360",
      transport: "http",
      url: "https://connect.mcp360.ai/v1/mcp360/mcp?token=YOUR_TOKEN",
    },
  ]}
>

Get your token at mcp360.ai.

MCP-UI: Interactive Components in Chat

Here's what makes this different: MCP servers can return UI, not just text.

Ask your copilot to "show product options" and get clickable cards. Run a poll. Display a chart. All inside the chat.

Examples:

Product cards with "Add to Cart" buttons
Voting polls
Data tables
Approval workflows
Form inputs

The server sends a UI schema, the SDK renders it. Your copilot becomes an interactive assistant, not just a text box.

Multiple Servers

Connect as many sources as you need:

mcpServers={[
  {
    name: "mcp360",
    transport: "http",
    url: "https://connect.mcp360.ai/v1/mcp360/mcp?token=YOUR_TOKEN",
  },
  {
    name: "linear",
    transport: "http",
    url: "https://mcp.linear.app",
    headers: { Authorization: `Bearer ${linearToken}` },
  },
  {
    name: "supabase",
    transport: "http",
    url: "https://mcp.supabase.com",
    headers: { Authorization: `Bearer ${supabaseToken}` },
  },
]}

Tools are namespaced automatically: mcp360_web_search, linear_create_issue, supabase_query.

What You Can Build

Use Case	Tools
Sales copilot	Lead enrichment, company research, CRM sync
Support assistant	Ticket lookup, knowledge base, escalation
Research tool	Web scraping, competitive analysis, data aggregation
Internal ops	Database queries, reports, workflow automation

Get Started

git clone https://github.com/YourGPT/copilot-sdk
cd copilot-sdk/examples/mcp-demo
npm install && npm run dev

View Live Demo →

Links: GitHub · MCP Docs · MCP360

Build Your Own AI Copilot with the Copilot SDK 💫

Rege — Wed, 04 Feb 2026 14:36:17 +0000

AI agents and chatbot systems have evolved so much that they're no longer just answering queries or handling escalations—they're now capable of managing almost all tasks themselves without any human intervention, especially in SaaS businesses and customer support.

Having a chatbot can get your customers' questions answered and automate some basic tasks. The better ones might even take actions and trigger workflows.

But here's the problem with most solutions out there:

They either just chat — answer questions, maybe search a knowledge base, then hand off to a human the moment things get real.

Or they're black boxes — limited customization, no flexibility, can't truly make it your own. You're stuck with what they give you.

If you want an AI Chatbot that actually takes action — not just chats — YourGPT Chatbot does that. Train it on your data, connect it to your tools, let it handle support and run workflows. Deploy in minutes.

But What If You Want to Build Your Own?

What if you want something that lives inside your product? Native. Seamless. Fully yours.

What if you need control over:

How it looks — Your UI, your design, your brand
Where it's deployed — Your servers, your cloud, your infrastructure
Where data is stored — Your database, your privacy rules, your compliance
What actions it takes — Your tools, your APIs, your logic

That's the gap. Most chatbot platforms don't let you build. They let you configure.

Copilot SDK: Build AI Copilots Your Way

We just released Copilot SDK — an open-source toolkit to create AI copilots for your app.

Design how it looks. Render any React component in chat — cards, forms, charts, buttons. The chat window becomes a canvas.

Define what it does. Create frontend tools (update UI, navigate, trigger modals) and backend tools (call APIs, update databases, send emails).

Pick any LLM. One line to switch between GPT-4, Claude, Gemini, or Grok.

Deploy anywhere. Self-host on your infrastructure. Your data stays yours.

Get Started

This is a quick sample — check the full documentation for complete setup and examples.

pnpm add @yourgpt/copilot-sdk @yourgpt/llm-sdk @anthropic-ai/sdk

import { createRuntime } from '@yourgpt/llm-sdk';
import { createOpenAI } from '@yourgpt/llm-sdk/openai';
....

const runtime = createRuntime({
  provider: createOpenAI({ apiKey: process.env.OPENAI_API_KEY }),
  model: 'gpt-5.1',
  systemPrompt: 'You are a helpful assistant.',
});

import { CopilotProvider, CopilotChat } from '@yourgpt/copilot-sdk';

<CopilotProvider>
  <CopilotChat />
</CopilotProvider>

That's it. You're live.

→ Docs
→ Examples
→ GitHub

Want quick deployment? → YourGPT Chatbot
Want to build your own? → Copilot SDK

Your AI. Your rules.

#ai #opensource #copilot #saas #react

MCP for Bulk Lead/Email Verification!✅

Rege — Sun, 26 Oct 2025 11:40:18 +0000

I had a list of 200 email addresses that needed verification. Before using them, I'd go through the usual process: export to CSV, upload to a verification service, wait for processing, download results, manually update my database, then re-import.

There had to be a simpler approach.

Now I just share the list in my AI chat and get verification results in seconds. No exports, no switching between tools, no interrupted workflow.

When verification is this straightforward, I'm more likely to maintain clean data consistently.

Why This Workflow Sucked

The traditional process looked like this:

Export leads from your CRM
Open some verification service
Upload the file
Wait (and wait)
Download results
Manually figure out what changed
Re-import everything

Every. Single. Time.

And if you had multiple batches? Multiply that pain. The context-switching alone killed productivity.

I thought: "My AI can analyze data, write code, and solve problems. Why can't it just... verify emails?"

Turns out it can. That's what MCPs do—they connect your AI to actual tools so you stay in the conversation.

How It Actually Works

Just share your list—CSV, pasted text, whatever*

Get instant breakdown: valid, invalid, risky emails all categorized

Example conversation:

You: "Here's a CSV with 200 conference leads, verify all emails"

AI: "Done! Found:

✅ 170 valid emails (85%)

❌ 18 invalid (9%)

⚠️ 10 risky catch-all domains (5%)

❓ 2 unknown

Created two lists: confirmed emails ready for outreach, and flagged ones for manual review."

That's it. No leaving the chat. No tool gymnastics.

Beyond Just "Valid or Invalid"

Once you're verifying in chat, you can do more interesting stuff:

Smart Segmentation

"Separate these by domain and tell me which companies have the most leads"

Batch Processing

"Here are 5 CSVs from different events, verify all and merge the valid ones"

Quality Analysis

"What percentage of our signup list is using temporary email services?"

Cleanup Automation

"Remove all invalid emails and update this list"

All the tedious stuff you'd manually do? Just ask.

What Changed For Me

Before:

😤 Export → Upload → Wait → Download → Update
⏱️ 15-20 minutes per batch
🤯 Context-switching hell
😑 "I'll verify these... later" (never happens)

After:

💬 Paste → Ask → Done
⚡ Results in seconds
🎯 Stay in flow
✅ Actually verify things now

The best part? I verify way more often because it's not annoying anymore. Better data quality, less bounced emails, happier inbox providers.

Want This Kind of Flow?

This is what MCPs do—connect your AI to tools so you don't have to leave the conversation. Email verification is just one example.

There are tons of MCPs out there for different things: data enrichment, web scraping, SEO audits, market research. They all work the same way—keep you in flow.

If you want to try email verification:

Check out mcp360.ai (has this plus 100+ other tools)
Or explore other MCP providers in the ecosystem

Setup is usually: grab a config line, paste it into your AI platform, done.

The bigger idea? AI gets way more useful when it can actually do things with your data, not just talk about it. MCPs make that happen.

One conversation. Less friction. More done.

More Tools & Resources

Visit mcp360.ai to explore other workflow automations—SEO, research, data tools, all accessible through your AI chat.

Use MCP to Automate Brainstorming Domain Names

Rege — Sun, 26 Oct 2025 11:39:24 +0000

You know that feeling when the perfect app name clicks in your head. It fits, it sounds right, and you want to see if the domain is free before you lose it.

You open a GoDaddy or Namecheap, wait for them to load, try the name, see it is taken, try another extension, still taken. Before long, you have ten tabs open and your focus is gone.

That used to happen all the time until I started using Claude with MCP360’s domain MCP. Now I just ask Claude to check availability and give me the link. It takes seconds.

The best part is that I do not break my thought process. When validation happens as fast as the idea itself, I can move from thinking to shipping without hesitation. That single change saves hours and keeps creative work alive.

How does it really help?

We've all been here: You're chatting with Claude about your new project—maybe it's an AI tool, a newsletter, a service, whatever. Ideas are flowing. Then you need to check if a domain is available, and suddenly you're context-switching like a maniac—leaving the chat, googling domain checkers, opening tabs, comparing prices, forgetting what you already checked.

With AI making it easier than ever to launch tools and services, more people are building and shipping. But the workflow? Still broken.

After doing this for the 100th time, I thought: "Why am I leaving the conversation to do something my AI could just... do?"

That's when I found MCP360. Instead of momentum killers, I can just ask Claude to check domains right there in the chat.

See It In Action

Natural conversation flow - just talking about your project idea

Instant validation without leaving the chat - see what's available, what's taken, and direct registration links

Beyond Basic Checks

Once you're in flow, you can do way more than just check if a domain is available:

WHOIS Lookup

"Who owns this domain?" Get registration dates, expiration info, registrar details—all without leaving the chat.

DNS Records

"Is my domain pointing to the right place?" Check DNS records when something's not working.

Bulk Checking

"Here's a CSV with 50 name ideas, tell me which are available." Perfect for those brainstorm-everything sessions.

All the boring stuff you'd normally juggle across multiple sites? Just ask.

What Changed For Me

Before:

🧠 Think of name
🛑 Stop everything
🌐 Open GoDaddy
🔍 Search domain
❌ It's taken
🔄 Repeat 10 times
😐 Return to AI with zero energy

After:

🧠💡 Think + Check + Decide + Move on
✅ Done

The best ideas happen when you're not thinking about process. Every time you leave that conversation to check a domain, you risk losing the idea itself.

Want This Kind of Flow?

This is what MCPs (Model Context Protocols) do—they connect your AI to actual tools so you can stay in the conversation. Domain checking is just one example.

There are tons of MCPs out there now: GitHub integration, database queries, file systems, web scraping, you name it. They all work the same way—keep you in flow instead of tab-switching.

If you want to try domain checking specifically:

Check out mcp360.ai (has 100+ tools including domains)
Or browse other MCP providers—there's a whole ecosystem growing

Setup is usually the same: grab a config line, paste it into your AI platform, done.

The bigger point? AI gets way more useful when it can actually do things, not just talk about them. MCPs make that happen.

One conversation. Less friction. More shipping.

More Tools & Resources

Check out mcp360.ai for 100+ other tools that keep you in flow—SEO analysis, web scraping, market research, and more.