Lab Ledger — MIT Critical Data

A single-page dashboard for tracking every paper in the lab — what stage it's at, which journal/conference it's currently with, and its full submission history (so a rejection-and-resubmission is visible, not overwritten).

No login, no build step, nothing to install to view it.

Structure

File	Purpose
index.html	The whole site. One file, no framework, no build step.
data.json	The real data the site reads, generated from the spreadsheet.
data.sample.json	Fixture data (8 papers) covering every UI state, for local testing.
export_data.py	Regenerates data.json from Lab_Papers_Dashboard.xlsx.
Lab_Papers_Dashboard.xlsx	Source of truth. Papers Tracker + Submissions Log + Team Directory.
assets/mit-critical-data-logo.svg, assets/favicon.ico	Brand assets, pulled from criticaldata.mit.edu.

Data flow

Lab_Papers_Dashboard.xlsx  →  export_data.py  →  data.json  →  index.html
     (people edit this)         (one command)      (site reads this)

The spreadsheet is where the lab enters information. The site never writes anything back — it's a read-only view generated from that spreadsheet.

Each paper can be sent to multiple journals/conferences over time. Instead of overwriting a paper's venue/deadline on every attempt, each attempt is logged separately in the spreadsheet's Submissions Log and exported into that paper's history array, so a rejection followed by a resubmission is fully visible on the paper's timeline rather than lost.

Updating the site after editing the spreadsheet

python3 export_data.py Lab_Papers_Dashboard.xlsx
git add data.json
git commit -m "Update paper statuses"
git push

GitHub Pages redeploys automatically within a minute or two of the push.

Deployment

Repo → Settings → Pages → Source: Deploy from a branch → Branch: main, folder / (root) → Save.

A private repo does not mean a private site — GitHub Pages URLs are publicly reachable by default even from a private repo. If that matters for this data, either add a password gate to index.html, or host on Vercel/Netlify instead, both of which support real password protection on free tiers.

Viewing this locally

Never open index.html by double-clicking it. That loads it as a file:// URL, and browsers block fetch() from reading local files under that scheme for security reasons — the page will silently render nothing but the header logo, with no on-screen explanation (the error only shows up in the browser console). Always serve the folder over HTTP instead:

cd lab-ledger
python3 -m http.server 8000
# or: npx serve

Then open http://localhost:8000/ in your browser. This is also exactly how GitHub Pages serves it in production, so testing this way matches reality.

To preview the dashboard with fake fixture data (data.sample.json, 8 papers covering every status) instead of the real data.json, open http://localhost:8000/?data=data.sample.json.