inkeep/agents
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.
What it solves
It provides a unified platform for building AI agents that bridges the gap between technical and non-technical team members. It allows users to create agents using either a visual drag-and-drop interface or a code-first TypeScript SDK, ensuring both methods stay synchronized in real-time.
How it works
The platform uses a multi-agent architecture where agents can be defined declaratively. It integrates with the Vercel AI SDK to interface with various LLM providers and supports Model Context Protocol (MCP) tools for extending agent capabilities. The system consists of an API for configuration and execution, a visual management UI, a TypeScript SDK for developers, a CLI for syncing code and visual configurations, and a UI component library for embedding chat interfaces.
Who it’s for
It is designed for engineering teams who want type-safety and CI/CD workflows, as well as non-technical teams (such as customer experience, sales, or marketing) who want to manage agent behavior without writing code.
Highlights
- Two-way Sync: Changes made in the Visual Builder are reflected in the TypeScript SDK and vice versa.
- Multi-agent Architecture: Supports teams of agents working together.
- MCP Integration: Includes MCP tools with built-in credential management.
- Observability: Provides a Traces UI and OpenTelemetry support for monitoring agent performance.
- Deployment Flexibility: Can be deployed via Vercel or Docker.