mirror of
https://github.com/n8n-io/self-hosted-ai-starter-kit.git
synced 2025-11-29 00:23:13 +00:00
115 lines
3.4 KiB
Markdown
115 lines
3.4 KiB
Markdown
# Self-hosted AI Starter Kit - Vision & Contribution Guidelines
|
|
|
|
Awesome that you're interested in contributing to the Self-hosted AI Starter
|
|
Kit! These specific guidelines are in addition to the general [n8n
|
|
contribution
|
|
guidelines](https://github.com/n8n-io/n8n/blob/master/CONTRIBUTING.md).
|
|
|
|
## Vision Statement
|
|
|
|
The Self-hosted AI Starter Kit is designed to be **the fastest path from zero
|
|
to working AI workflows** for developers and organizations who want to
|
|
experiment with local, private AI solutions. It provides a curated,
|
|
pre-configured foundation that "just works" out of the box, enabling users to
|
|
focus on building AI workflows rather than wrestling with infrastructure
|
|
setup.
|
|
|
|
## Core Principles
|
|
|
|
### 1. Simplicity Over Completeness
|
|
|
|
The starter kit should prioritize ease of use and quick setup over
|
|
comprehensive feature coverage. It's better to do fewer things well than to
|
|
attempt to solve every possible use case.
|
|
|
|
### 2. Learning-Focused, Not Production-Ready
|
|
|
|
This is explicitly a **learning and experimentation platform**. Users should
|
|
be able to go from `git clone` to working AI workflows in minutes, not hours.
|
|
Production-grade concerns like high availability, advanced security, and
|
|
scalability are intentionally out of scope.
|
|
|
|
### 3. Opinionated but Extensible
|
|
|
|
We make opinionated choices about the core stack (n8n + Ollama + Qdrant +
|
|
PostgreSQL) to reduce decision paralysis, while providing clear paths for
|
|
users to extend and customize as they learn.
|
|
|
|
### 4. Privacy-First Local Development
|
|
|
|
Everything should work completely offline and locally by default. External
|
|
dependencies should be minimal and optional.
|
|
|
|
## What Belongs in the Starter Kit
|
|
|
|
### Core Components
|
|
|
|
- **n8n**: The workflow automation platform
|
|
- **Ollama**: Local LLM inference
|
|
- **Qdrant**: Vector database for embeddings
|
|
- **PostgreSQL**: Persistent data storage
|
|
- **Basic networking**: Simple Docker networking to connect components
|
|
|
|
### Essential Configuration
|
|
|
|
This includes:
|
|
- Pre-configured environment variables with sensible defaults
|
|
- Basic Docker Compose profiles for different hardware (CPU, GPU-Nvidia, GPU-AMD)
|
|
- Minimal volume mounts for data persistence
|
|
- Sample workflow demonstrating the core capabilities
|
|
|
|
### Getting Started Materials
|
|
|
|
This includes:
|
|
- Clear installation instructions for different platforms
|
|
- A demo workflow showcasing AI capabilities
|
|
- Basic documentation for accessing local files
|
|
- Links to relevant n8n documentation and templates
|
|
|
|
## What Doesn't Belong in the Starter Kit
|
|
|
|
### Production Infrastructure
|
|
|
|
Including:
|
|
- Reverse proxies
|
|
- SSL/TLS termination
|
|
- Load balancers
|
|
- Advanced monitoring and logging
|
|
- Backup and recovery systems
|
|
- Container orchestration beyond basic Docker Compose
|
|
|
|
### Advanced Networking
|
|
|
|
Including:
|
|
- Custom network configurations
|
|
- VPN integrations
|
|
- Multiple environment setups
|
|
- Advanced security hardening
|
|
|
|
### Alternative Technology Stacks
|
|
|
|
Including:
|
|
- Different vector databases
|
|
- Alternative workflow platforms
|
|
- Multiple LLM backends beyond Ollama
|
|
- Different databases for the core setup
|
|
|
|
### Enterprise Features
|
|
|
|
Including:
|
|
- Authentication systems
|
|
- Multi-tenancy
|
|
- Advanced access controls
|
|
- Compliance tooling
|
|
|
|
## PR specific requirements
|
|
|
|
- Small PRs Only:
|
|
- Focus on a single feature or fix per PR.
|
|
- Typo-Only PRs:
|
|
- Typos are not sufficient justification for a PR and will be rejected.
|
|
|
|
|
|
Remember: **It's better to be an excellent starting point than a mediocre
|
|
everything-solution.**
|