Hexabot: Build AI Workflows for Conversational Channels

Overview

Hexabot brings workflow automation and conversational AI into a shared runtime. Teams can define agentic workflows and actions, then connect them to conversational channels, with support for tools, memory, MCP integration points, and RAG.

It is aimed at developers building AI-powered bots or automation services that need to coordinate tasks across channels. A CLI-generated project provides a starting point; the repository also contains the monorepo for teams contributing to the platform itself.

Key Features

• YAML workflow definitions with typed runtime contracts.
• Actions that describe workflow behavior, with schema-validated inputs, outputs, and settings.
• Reusable bindings that separate capabilities and configuration from task logic.
• Runtime support for explicit memory definitions.
• MCP integration points for tool and context interoperability.
• Conversational channels and helpers for multi-channel applications.
• TypeORM data access, with SQLite as the local default and PostgreSQL supported for production.

Use Cases

• A product team can build a conversational assistant that runs defined workflows and exposes it through its chosen channels.
• An automation developer can combine reusable actions and bindings to handle recurring tasks with AI capabilities.
• A team exploring MCP can use the platform's integration points to connect workflow execution with tools and context.
• A team prototyping an AI bot can generate a project with the CLI and run its local admin UI and API.

Getting Started

Install the CLI and create a project:

npm install -g @hexabot-ai/cli
hexabot create my-project
cd my-project
hexabot dev

The CLI requires an interactive terminal when creating a project. See the README and documentation for setup details.

License in Plain Words

No standard license was detected. Without a license, the default copyright rules apply: you may read the code, but not reuse it. Check the repository or ask the authors.

Alternatives

• magic: Magic focuses on managed AI automation with organizational tools, approvals, and sandboxing, rather than running conversational bots across channels.
• LibreChat: LibreChat centers on a self-hosted, multi-provider chat interface with agents and tools, rather than a configurable workflow runtime for conversational channels.

Considerations

The project is marked active, with regular recent commits and substantial merged pull-request activity. However, it has had no release in over a year, so teams should verify that the current main branch meets their stability and upgrade expectations. Contributions are spread across multiple contributors rather than depending mainly on one person.

The supplied project checks report no detected test suite. The documented runtime requires Node.js ^24.17.0; database configuration uses TypeORM, with SQLite for local use and PostgreSQL available for production. Docker is optional for Docker-based services. The README identifies the license as FCL-1.0-ALv2, while repository metadata reports NOASSERTION, so review the repository's license file and obligations before adopting it.