OSS / About
Who We Are
The team and community behind Aurora Gateway — what we are, how we work, and where to find us.
Browse docs
What Aurora is
You can run Aurora on a laptop, in Docker, on Kubernetes, or behind a corporate proxy. The same aurora binary serves local development, a single-team gateway, and a multi-tenant production deployment. Configuration is layered (code defaults, config.yaml, environment variables) so the same artifact behaves differently across environments without rebuilds.
Who maintains it
Aurora is maintained in the open. The team works on the public repository, accepts pull requests, and ships releases under Conventional Commits. The implementation is intentionally small and explicit — large abstractions, magic, and framework dependencies are deliberately avoided.
Contributions of every kind are welcome: bug reports, documentation fixes, provider integrations, guardrail recipes, and dashboard work. See Contributing for the workflow and Architecture for the codebase layout.
Principles
These are the rules the project operates by. They show up in every design decision and code review.
Generous on the input, conservative on the output
Accept OpenAI- and Anthropic-format requests as-is, even when they include non-standard fields. Translate them to whatever the upstream provider actually needs. Return responses in a clean OpenAI or Anthropic shape regardless of what the upstream returned. This is the Technical Philosophy made concrete.
Good defaults beat mandatory configuration
A user who has set OPENAI_API_KEY and started the binary should have a working gateway. Every config flag exists because someone asked for it; nothing is required unless the alternative is unsafe.
Operational visibility is a feature
Logs, metrics, audit records, the live console stream, and the dashboard are not afterthoughts. They are how operators know what the gateway is doing. When something goes wrong, the answer is in the data.
Small, explicit, easy to operate
The binary fits on a thumb drive. Configuration is plain YAML or env vars. Dependencies are pinned. Builds are reproducible. Config, logs, and process management follow the same disciplined patterns — no surprises, no mystery state.
Where to find us
GitHub
Source code, releases, issues, and pull requests.
Open link →
Discord
Real-time help, design discussions, and community announcements.
Open link →
npm
Install the aurora binary with npm install -g iaurora.
Open link →
Docker Hub
Pull aurorahq/aurora for containerized deployments.
Open link →
How to get involved
Start with the Quick Start. Send one request through the gateway, then explore Concepts to see what the rest of the system is.
Related
- Our Values — the three things we optimize for
- Technical Philosophy — the engineering ideas that shape the code
- Roadmap — what is shipping next
- Anonymous Heartbeat Telemetry — what Aurora reports back to the project