Agents

v0.73.5Coding Agentsstable

Create AI Agents in a No-Code Visual Builder or TypeScript SDK with full 2-way sync. For shipping AI assistants and multi-agent AI workflows.

agentagent-builderagent-frameworkagents-sdkai
Share:
1,153
Stars
0
Downloads
0
Weekly
0/5

What is Agents?

Agents is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to create ai agents in a no-code visual builder or typescript sdk with full 2-way sync. for shipping ai assistants and multi-agent ai workflows.

Create AI Agents in a No-Code Visual Builder or TypeScript SDK with full 2-way sync. For shipping AI assistants and multi-agent AI workflows.

This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Create AI Agents in a No-Code Visual Builder or TypeScript S

Use Cases

No-code visual agent builder
TypeScript SDK for AI assistants and workflows
inkeep

Maintainer

LicenseNOASSERTION
Languagetypescript
Versionv0.73.5
UpdatedMay 21, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

NPM

npx -y @inkeep/agents-mcp

Manual Installation

npx -y @inkeep/agents-mcp

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Agents

Inkeep Agents is a platform for building and deploying AI assistants and multi-agent workflows through a no-code visual builder and a TypeScript SDK, with full two-way sync between both authoring modes. The platform exposes MCP tools with built-in credential management, supports invocation via MCP, Agent-to-Agent (A2A) protocols, and the Vercel AI SDK, and includes observability through a Traces UI and OpenTelemetry integration. Teams shipping AI-powered support assistants, documentation bots, or complex multi-agent pipelines use it to define sub-agents with specific capabilities, connect them to tools, and monitor execution without managing infrastructure.

Prerequisites

  • Node.js 22 or later installed
  • pnpm 10 or later (the platform's package manager)
  • Docker installed and running (required for local development environment)
  • An Inkeep account and API credentials
  • An MCP-compatible client such as Claude Desktop for MCP integration
1

Create a new Agents project

Use the Inkeep create-agents scaffolding tool to generate a new project with the TypeScript SDK, Docker configuration, and example agent definitions.

npx @inkeep/create-agents my-agents
cd my-agents
2

Configure environment credentials

Copy the example environment file and populate it with your Inkeep credentials. The .env file is used by both the local dev server and Docker.

cp .env.example .env
# Edit .env and set:
# INKEEP_AGENTS_MANAGE_UI_USERNAME=your-username
# INKEEP_AGENTS_MANAGE_UI_PASSWORD=your-password
3

Start the local development environment

Run the setup and dev commands to initialize Docker services and launch the local Visual Builder at http://localhost:3000.

pnpm setup-dev
pnpm dev
4

Define agents using the TypeScript SDK

Create agent definitions using the subAgent helper. Each agent has an id, name, description, a list of MCP tools it can use, and a prompt defining its behavior.

import { subAgent } from '@inkeep/agents-sdk';

const supportAgent = subAgent({
  id: "support-agent",
  name: "Support Agent",
  description: "Answers customer questions using documentation",
  canUse: () => [documentationMcp],
  prompt: `You are a helpful support agent. Use the documentation tool to answer questions accurately.`
});
5

Add the Agents MCP server to your client

Configure your MCP client to connect to the Inkeep Agents platform, which exposes your defined agents as invocable MCP tools with managed credentials.

Agents Examples

Client configuration

Add the Inkeep Agents MCP server to claude_desktop_config.json using the official npm package.

{
  "mcpServers": {
    "inkeep-agents": {
      "command": "npx",
      "args": ["-y", "@inkeep/agents-mcp"],
      "env": {
        "INKEEP_AGENTS_MANAGE_UI_USERNAME": "your-username",
        "INKEEP_AGENTS_MANAGE_UI_PASSWORD": "your-password"
      }
    }
  }
}

Prompts to try

Example prompts once your agents are defined and the MCP server is connected.

- "List the available agents and describe what each one can do"
- "Invoke the support-agent with the question: how do I reset my password?"
- "Run the data-analysis agent on this CSV and summarize the key trends"
- "Show me the OpenTelemetry trace for the last agent run and identify any slow tool calls"
- "Create a new sub-agent that answers questions about our API documentation"

Troubleshooting Agents

pnpm dev fails because Docker is not running

Start Docker Desktop (macOS/Windows) or the Docker daemon (Linux: sudo systemctl start docker) before running pnpm setup-dev or pnpm dev. The local development environment requires Docker for its backing services.

The Visual Builder at localhost:3000 returns a 401 Unauthorized error

Verify that INKEEP_AGENTS_MANAGE_UI_USERNAME and INKEEP_AGENTS_MANAGE_UI_PASSWORD are set correctly in your .env file. Restart the dev server after editing .env: stop with Ctrl+C and run pnpm dev again.

npx @inkeep/agents-mcp fails to start or returns a connection error

Ensure Node.js 22+ is installed (check with `node --version`). Verify your credentials are correct and that the Inkeep platform endpoint is reachable. Check docs.inkeep.com for the current MCP server configuration requirements.

Frequently Asked Questions about Agents

What is Agents?

Agents is a Model Context Protocol (MCP) server that create ai agents in a no-code visual builder or typescript sdk with full 2-way sync. for shipping ai assistants and multi-agent ai workflows. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Agents?

Install via npm with the command: npx -y @inkeep/agents-mcp. Then add the server configuration to your AI client's JSON config file (e.g., claude_desktop_config.json or .cursor/mcp.json).

Which AI clients work with Agents?

Agents works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Agents free to use?

Yes, Agents is open source and available under the NOASSERTION license. You can use it freely in both personal and commercial projects.

Browse More Coding Agents MCP Servers

Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "agents": { "command": "npx", "args": ["-y", "@inkeep/agents-mcp"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Agents?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides