Load your scraper output into the Meetings Viewer and start reviewing in under a minute.

Quick start

Prerequisites: Node.js (>=20.11.0 - .nvmrc pins 24) and a one-time npm install in this repo.

1. Run your spider

From the city-scrapers repo, run your spider with output redirected to the viewer's data directory:

scrapy crawl <spider-name> -O data/scrapers/<spider-name>.json

Use uppercase -O, not lowercase -o. The lowercase flag appends to an existing file, which silently duplicates records. The uppercase flag overwrites, which is what you want.

2. Open the viewer

Start the Next.js dev server and navigate to the scrapers page:

npm run dev

Then open http://localhost:3000/scrapers in your browser. You'll see a list of all spiders with JSON output in data/scrapers/.

3. Inspect the output

Click your spider's name to see its output as a filterable table. Key things to check:

Duplicates are highlighted with a colored badge. Records with the same start time and a similar title are grouped.
Status is shown as a chip. Make sure cancelled and tentative meetings are marked correctly.
Location should not be empty unless time_notes explains why.
Links should be in priority order (agenda first, then minutes, then video, then other).

4. Compare against the rubric

Open the QA rubric in a second tab and check each field against the criteria. The rubric is designed to be applied field by field against real output.

5. Open a PR

Once your output passes the rubric, commit the JSON file and open a PR. Use the QA checklist (coming soon) to generate a PR comment with your review notes.

Quick Start