Skip to content

Genesis

Genesis is a cross-disciplinary researcher-debate engine.

Code Demo Video Cerebral Valley

Core trick

Agents grounded in each researcher's OpenAlex papers via RAG debate over a Zep temporal knowledge graph, then a novelty check queries OpenAlex.

Tags

scienceagents

Description

Genesis is a cross-disciplinary researcher-debate engine. You paste a research question, select real researchers from OpenAlex (no API key required), and Genesis convenes a live multi-agent debate: each agent is grounded in that researcher's actual papers (RAG). The agents debate across disciplines — biology, AI, cosmology, philosophy — over a Zep temporal knowledge graph where every claim, rebuttal, and bridge is a real edge. Claude Opus 4.8 then synthesizes the debate into a grounded, novel, testable hypothesis brief, with every claim traced back to a DOI. A Novelty Audit queries OpenAlex to verify whether the cross-disciplinary bridge is genuinely new (NOVEL / PARTIAL / KNOWN verdict). A D3 graph visualization glows each researcher's node by how much their ideas influenced the final hypothesis — clicking reveals their paper and DOI. A second "Debate Network" view shows who endorsed, challenged, and amplified whom across disciplines.
seyeong-han/Genesis★ 10 forksPythonAGPL-3.0Created Jun 13, 2026 · last push Jun 14, 2026

README excerpt

# Genesis — Cross-Disciplinary Research Debate Engine

> Convene a live panel of real researchers over their *actual* papers. Watch them
> debate your question, discover novel cross-disciplinary hypotheses, and see exactly
> whose paper sparked each idea — with DOIs you can verify.

Genesis does not assert answers. It surfaces **grounded, novel, testable leads**:
- **Grounded**: every claim cites a real paper (DOI)
- **Novel**: an OpenAlex Novelty Audit checks whether the bridge is already published
- **Testable**: the brief ends with a falsifiable next experiment

> Genesis is an AGPL-3.0 fork of [MiroFish](https://github.com/666ghj/MiroFish).
> MiroFish's three core ideas are inherited (shared temporal graph / stigmergic
> round accumulation / post-hoc synthesis) and the social-simulation semantics are
> replaced with a researcher-debate pipeline.

---

## Quick start

```bash
# 1) Copy and fill env variables
cp .env.example .env
#    Fill ANTHROPIC_API_KEY and ZEP_API_KEY

# 2) Import real researcher papers (one per discipline, 43 disciplines)
python backend/scripts/import_openalex_corpus.py --auto
# Note the project_id printed

# 3) Start backend + frontend
docker-compose up       # OR:
cd backend && python run.py   # backend :5001
cd frontend && npm run dev    # frontend :3000

# 4) Open http://localhost:3000
#    Use the project_id from step 2 to build the graph
```

## How it works (5-step pipeline)

1. **Graph Build** — OpenAlex papers ingested → Zep temporal knowledg

Team