Show HN: Heku – Dynamic MCP Tooling via JSON Configs
Heku is a tool that enables dynamic Model Context Protocol (MCP) tooling through JSON configuration files, allowing developers to define and manage MCP tools without writing code. It provides a config-driven approach to building and extending AI tool ecosystems.
Background
- **MCP (Model Context Protocol)** is an open standard by Anthropic that lets AI models (like Claude) interact with external tools, files, and services in a structured way. It's a key part of making LLMs agentic — able to take actions rather than just generate text.
- **HeKU** is a new tool that lets developers define MCP-based tools using simple JSON config files, without needing to write any code. This makes it much easier to connect AI models to custom APIs, databases, or internal systems.
- **Dynamic MCP tooling** means tools can be created, updated, or swapped at runtime — useful for teams who want to quickly prototype or iterate on AI agent behaviors without redeploying services.
- The project lives on **GitHub under RapidThoughtLabs** and targets developers building AI agents or copilot-style experiences who need a lightweight, config-driven alternative to building custom MCP servers from scratch.