GitHub Developer Portal
Take a deep dive under the hood of our agricultural research operating system. Clean architecture, robust orchestrator pipelines, and beautiful React UI.
Stack Architecture
Next.js 15, React, Firebase Firestore, Tailwind CSS, Google Gemini 1.5
Orchestrated Sources
Parallel fan-out queries spanning 17 concurrent citation indexing nodes
Open For Contributions
Pull requests for new providers (ServiceFactory, Zod Schemas) always welcome
The Orchestrator Architecture
The core engine of CatalystLab operates on a customized pipeline dividing workloads into 3 layered phases:
Retrieval & Validation
Client issues a concept collision. Parallel requests are spawned to index nodes. Every payload raw result is piped through Zod schema check blocks (e.g. ResearchResultSchema) before admission.
Curated Filtering
Applies the strict 3-year threshold limit. All papers outside 2023–2026 are purged. Surviving items are grouped by provider, sorted by relevance score, taking exactly the top 3 items per provider.
Dynamic Synthesis
Constructs a consolidated scientific context. Passes data block to Google Gemini endpoints, which performs semantic parsing of experimental variable bounds defined in the Collider.
Local Setup Guide
Clone, configure environment variables, and build the agricultural research lab locally in a matter of minutes.
Clone Repository
Download the source files from the official CatalystLab repository.
Install Dependencies
Install the packages from package.json in your local node environment.
Environment Configuration
Set up your server secrets. Open .env and add your Gemini API key.
Compile and Execute
Run the development server natively or build the full production client bundle.