CodeZero is an open-source, self-hostable platform for building automations, integrations, and backend logic as visual flows on a canvas, aimed at teams that want AI-assisted no-code workflow building while keeping their data on infrastructure they control.

What it is

CodeZero is a no-code automation and workflow platform where logic is assembled as a graph of nodes on a visual canvas rather than written as source code. A flow can be an automation, an integration, or an entire piece of backend logic: a trigger is connected, logic is chained node by node, and the flow is run. Triggers come built in as HTTP/webhook and cron, and plugins extend the runtime with additional nodes, flow types, and triggers. The editor is assisted by AI, which generates a flow from a natural-language prompt or a smart template and can subsequently refine, extend, or rewrite any part of the graph directly on the canvas.

The concrete problem it solves is boilerplate. Instead of writing glue code to connect two services, exposing an internal HTTP endpoint, or scheduling a recurring job, the user models the same behaviour as a flow. CodeZero lives in the low-code and self-hosted automation ecosystem, and its own topic list places it explicitly as an n8n-alternative. It also carries the mcp topic, indicating participation in the Model Context Protocol tooling space. In short, it replaces hand-written integration and backend plumbing, and it is positioned as a substitute for hosted workflow tools that keep execution and data on the vendor's side.

Key capabilities

• Visual flow editor that renders logic as a graph of nodes, intended to stay readable, debuggable, and shareable across a team.
• AI-assisted building: describe the desired automation or select a smart template, and CodeZero generates the flow; any part of an existing flow can be rewritten with a single prompt on the canvas.
• Bring your own model: connect OpenAI, OpenRouter, or any compatible provider, and choose which models power each workspace.
• Built-in HTTP/webhook and cron triggers, with plugins adding further trigger types.
• Expose any flow as an HTTP endpoint, allowing CodeZero to serve as a visual backend.
• Deployment freedom: fully self-hostable, or hybrid with the IDE in the cloud and runtimes on your own hardware, so data never leaves your network.
• Team structures: organizations, projects, roles, role templates, and fine-grained permissions; every runtime component scales independently from a single machine to a fleet.

Who uses it and how

• Teams automating service-to-service integrations, for example posting a new GitHub issue to Discord or creating a DHL shipment from a new Shopify order.
• Developers who need an internal or public HTTP API without writing backend boilerplate, by exposing a flow as an endpoint.
• Operations groups running scheduled work on cron for reports, syncs, and cleanups.
• Teams modelling multi-step business processes where branching, data control, and error handling must be explicit.
• Organizations that need segregated projects, role-based access, and optionally a hybrid split between a cloud IDE and self-hosted runtimes.

Getting started

Docker is the only prerequisite. Clone https://github.com/code0-tech/codezero.git, move into the docker-compose directory, adjust secrets and settings such as ports, TLS, COMPOSE_PROFILES, and available AI models in .env, then run docker compose up -d and open http://localhost.

How it compares

The registry facts name no paid products that CodeZero replaces, but its own topics place it against hosted workflow automation tools such as n8n, in the same self-hosted, low-code lane. The contrast the project itself draws is deployment and ownership: self-hosting or hybrid execution keeps data inside the operator's network, and the runtime can scale component by component rather than as one managed service. Cost and data-ownership decisions therefore sit with the operator rather than with a vendor subscription.

When to use it — and when not to

A self-hoster must operate a Docker host and manage the .env configuration, including secrets, ports, TLS, and model-provider credentials; the README explicitly warns that the default credentials ([email protected] / root) and default secrets must be changed before anything is exposed publicly. The licence is reported as NOASSERTION, so the terms are unclear and should be verified before commercial reliance. The project is also small — 55 stars, 3 forks, and 9 open issues — so anyone needing a vendor SLA, a large plugin ecosystem, or clearly stated licensing should look elsewhere or proceed only after reviewing the repository.

git clone https://github.com/code0-tech/codezero.git
cd codezero/docker-compose

# adjust secrets & settings (ports, AI models, ...) in .env, then:
docker compose up -d