ARA - DeSci platform

Independent researchers and prolific hackathon builders, like myself , often produce valuable work that ends up as "project presentations" without a formal, citable home. The Amateur Research Association (ARA) is a self-publishing system designed to solve this. It's an automated pipeline that provides a legitimate, low-cost, and professional-looking venue for scholarly work.

The system is designed to take LaTeX manuscripts and metadata, and automatically publish them as a cohesive, static website hosted on GitHub Pages. The final site includes a main landing page with a table of contents , individual metadata pages for each article , and links to the full-text PDF downloads.

For this hackathon, we are demonstrating the foundational proof-of-concept. While the automation pipeline is not yet fully polished, it successfully demonstrates the complete, end-to-end flow: from a set of LaTeX files in a repository to a live, published release. This project (detailed in its own self-published whitepaper) lays the groundwork for a scalable, community-owned scholarly commons where all contributions can be formally recognized.

It's slightly out of scope / underdone for this event I guess, at least until it has more integrations. But I think this has potential (and could be integrated for serving ecosystem needs once I gather some data from potential users over the next month) and I'll probably build it further.

This project is a fully open-source, automated pipeline built entirely around GitHub.

Content & Format: Authors write manuscripts in LaTeX using the standard IEEEtran.cls class file to ensure a professional, uniform appearance. All metadata is managed in simple YAML files. (Yeah, using Markdown and templating was considered. In a more serious project it will be necessary.)

Automation: The core of the project is a GitHub Actions workflow. When triggered, it sets up a LaTeX environment and runs a custom Python build script.

Build Process: A Python script, using Jinja2 templates (like release.html.j2 ), reads all the submission metadata, assembles the front matter and individual articles and generates PDFs and the static HTML for the website.

Compiling & Deployment: The script calls pdflatex to compile all LaTeX sources into PDFs. The final static site (HTML and all PDFs) is then automatically deployed and "hosted permanently on GitHub Pages".

Proof-of-Concept Status: The automation is functional but basic. For example, we initially planned to use the towncrier utility for release assembly but ultimately "decided not to implement that for now" to focus on demonstrating the core end-to-end functionality for the hackathon. Same with provenance, while it will be essential to provide transparency between initial authorship and final artifacts that goes into advanced category that needs some R&D. There are few interesting ideas for potential tokenomics but I consider that secondary and didn't implement any for now.