Publish WG Minutes

Overview

This skill covers processing WebPerfWG meeting minutes exported as zip files from Google Docs and publishing them to the web-performance repo.

Working Directory

All work happens in the meetings directory of the web-performance repo: /Users/yoavweiss/OS/web-performance/meetings/

Input Format

Zip files are named like:

• WebPerfWG call - December 18th, 2025.zip
• WebPerfWG call - October 9th 2025.zip

Each zip contains:

• An HTML file (e.g., WebPerfWGcallDecember18th2025.html)
• An images/ directory (optional, contains referenced PNGs)

Process

1. Create the target directory

Use the pattern {year}/{year}-{MM}-{DD}/ with zero-padded month and day:

2025/2025-12-18/
2025/2025-10-09/
2026/2026-01-15/

2. Copy, unzip, and clean up the zip

cp "WebPerfWG call - December 18th, 2025.zip" 2025/2025-12-18/
cd 2025/2025-12-18 && unzip *.zip && rm *.zip && cd ../..

3. Rename the HTML to index.html

mv 2025/2025-12-18/WebPerfWGcallDecember18th2025.html 2025/2025-12-18/index.html

4. Run cleaner.sh

The cleaner.sh script in the meetings directory applies formatting fixes (bold speaker names, remove Google Docs boilerplate, fix diacritics like Nicolás Peña, adjust layout):

bash cleaner.sh 2025/2025-12-18/index.html

5. Delete the original zip file

rm "WebPerfWG call - December 18th, 2025.zip"

6. Update README.md

Add links under the appropriate year heading in README.md, in reverse chronological order. The link format is:

* [December 18th](https://w3c.github.io/web-performance/meetings/2025/2025-12-18/index.html)

If a new year section is needed, add it above the previous year's heading:

# 2026

* [January 15th](https://w3c.github.io/web-performance/meetings/2026/2026-01-15/index.html)

# 2025

Date Parsing

Extract the month, day, and year from the zip filename. Map month names to numbers:

• January=01, February=02, ..., December=12

Strip ordinal suffixes (st, nd, rd, th) from the day and zero-pad to 2 digits.

Checklist

• Directory created with {year}/{year}-{MM}-{DD}/ pattern
• Zip extracted and removed
• HTML renamed to index.html
• cleaner.sh run on the HTML
• Original zip deleted from meetings root
• README.md updated with link in correct chronological position