Build, run, and operate AI agents, workflows, skills, knowledge, and model routes from one workspace.
Why ZGI · Platform · Quick Start · Architecture · Development · Docs · License
ZGI is a source-available AI application platform for teams that need more than a chat box. It brings agent apps, visual workflow orchestration, model operations, knowledge retrieval, reusable skills, and runtime services into one self-hostable product surface.
Use it to prototype internal AI tools, publish agent experiences, route model traffic through controlled providers, compose multi-step automations, and keep runtime infrastructure close to your own workspace.
| Area | What you can build |
|---|---|
| Agent applications | Configure instructions, model settings, knowledge, memory, file upload, skills, and web app publishing from the console. |
| Visual workflows | Compose API calls, JSON parsing, LLM steps, branching, loops, approvals, tools, code execution, database access, notifications, and retrieval on a canvas. |
| Model operations | Manage providers, channels, credentials, defaults, policy controls, and pricing metadata without scattering model configuration across apps. |
| Knowledge and skills | Connect datasets, content parsing, retrieval, and reusable tool skills so agents can work with real workspace context. |
| Runtime services | Run the Go API, Next.js console, sandbox, plugin runner, PostgreSQL, and Redis behind a local gateway at http://localhost:2679. |
| Starter templates | Explore 23 built-in agent and workflow templates in English and Simplified Chinese. |
Start the full local stack:
make dev-dockerIf you do not have make, run the startup script directly:
./dev/start-dockerOpen the console:
http://localhost:2679
On first launch, create the first administrator account. ZGI does not ship with a default admin account.
| Command | Services |
|---|---|
./dev/start-docker --core |
Core product preview |
./dev/start-docker --runtime |
Core stack plus Sandbox and Runner |
./dev/start-docker --knowledge |
Core stack plus knowledge dependencies |
./dev/start-docker --full |
Full stack, same as the default |
Stop the stack:
make docker-downView logs:
make docker-logsBrowser
|
v
Nginx gateway (:2679)
|
+-- Next.js console (web)
+-- Go API service
|
+-- PostgreSQL / Redis
+-- Sandbox service
+-- Plugin runner
+-- Optional knowledge services: Weaviate / Neo4j
.
+-- api/ Go backend service
+-- web/ Next.js web console
+-- sandbox/ Isolated execution service
+-- runner/ Plugin execution service
+-- docker/ Product-level Docker Compose assets
+-- dev/ Local development scripts
+-- scripts/ Maintenance scripts
+-- docs/ Public documentation and assets
+-- Makefile Common local entry points
+-- README.md
Install the local toolchain first:
- Docker and Docker Compose
- Make
- Go, for backend source development
- Node.js and pnpm, for frontend source development
The web app uses pnpm@10.12.1.
Install and prepare dependencies:
make setupStart shared infrastructure:
make dev-dockerRun backend and frontend from source in separate terminals:
make dev-api
make dev-web- Product-level Docker notes:
docker/README.md - Release process:
docs/release-process.md - Script skill input files:
docs/script-skill-input-files.md - Web app notes:
web/README.md - Backend service docs:
api/
Contributions are welcome. Please read CONTRIBUTING.md
before opening a pull request.
Community expectations are documented in
CODE_OF_CONDUCT.md.
For security-sensitive reports, follow SECURITY.md.
ZGI is source-available under the ZGI Community License, based on Apache License
2.0 with additional conditions. ZGI is free for personal, research,
educational, and internal organizational use. Hosted multi-tenant services,
white-label distribution, and removal of official ZGI branding require a
commercial license. See LICENSE for details.
The Apache License 2.0 text referenced by the ZGI Community License is included
in LICENSE-APACHE.
