Science needs more than copilots.
A research platform that takes an open scientific question through a controllable, automated run and produces a manuscript, code, and reviewable artifacts you can execute again. Not a chat that writes about research — a workflow whose output is a package that can be challenged and re-run.
omegaXiv
In developmentIn development, and producing real output. The run engine, paper workspace and knowledge graph are shipped; several integrations remain deliberate seams that run hermetically until a key is configured.
Python · FastAPI · Next.js · PostgreSQL · Redis · Terraform
The gap omegaXiv addresses is not generation. Models already write plausible scientific prose. The gap is that a generated answer arrives detached from everything that would let somebody check it — the code that produced it, the artifacts it depends on, the assumptions it ran under, and the ability to run it again and get the same thing.
So the unit of work here is not a response. It is a run: a bounded execution with a brief, a runtime, a state machine, an artifact bundle, and a review surface, which exports as a package somebody else can open.
What it has actually produced
| Measurand | Value | Method | Uncertainty |
|---|---|---|---|
| Published research runs | 45 | public repositories under the alphaomega-labs organization | ±0, countable |
| Contents of each published run | 8 artifacts | paper (LaTeX + PDF), sources, experiments, knowledge, manifest, artifacts, solution, review | ±0, inspectable |
| Peer review of the output | none | no external review process exists | unsupported — these are machine-produced drafts, not accepted papers |
How a run works
Structured intake
A problem page carries the brief, the runtime assumptions, and the launch controls, so new work starts from structured context rather than an empty prompt.
- Structured problem brief
- Launch controls
- Shared execution context
The quarks run engine
Dispatch and a phase runtime — render, turn, parse, validate, then repair or retry — under a control plane with an explicit run state machine.
- Phase runtime with validation
- Repair and retry
- Explicit run state machine
Private run workspace
Pipeline state, quality signals, cost, and re-run actions stay together while a question is still being worked through.
- Run detail and progress
- Quality and cost tracking
- Private execution controls
Paper, code, and artifacts
A publication workspace keeps the manuscript, the artifact bundle, and the review surface attached to each other, so the output stays legible after the run ends.
- Paper IDE with live preview
- Artifact links and previews
- Review-ready package
The Meta Controller
An autonomous interrupt supervisor that polls for stalled runs, decides whether to intervene, and applies a typed intervention or escalates to a person. Off by default.
- Detects needs-input, blocked, paused
- Typed interventions
- Escalation to a human
Research memory
A per-run artifact graph plus a global cross-run store of entities and relations, so lineage survives and each pass can build on the last rather than resetting.
- Per-run artifact graph
- Cross-run entity store
- Retained lineage
The workspace
Problem detail: the brief, current status, run history, and launch configuration all stay attached to the same research object.Open full size ↗ 
Run detail: pipeline progress, quality tracking, cost visibility, and next actions in one private workspace while the research is active.Open full size ↗ 
The paper workspace keeps the manuscript, artifact bundle, source problem, and publication controls together.Open full size ↗ 
The discovery feed surfaces active research requests, tags, and review context so work can be found and evaluated in public.Open full size ↗
Where it sits
omegaXiv is the research prototype; Cortex is what closes the gap from a reviewable result to something running in production. The two share a thesis — that the apparatus around a claim matters as much as the claim — and Cortex inherited the governed-execution machinery that omegaXiv needed first.
Go and look
- Published research runsEvery run exports as its own public repository: paper, sources, experiments, review, manifest
- DWG
- A2O-000
- Sheet
- 04 of 09
- Rev
- draft
- Scale
- 1:1