Add automated bibliography system for newsletter using Zotero/BibTeX

[Copilot]

Description

Implements automated bibliography management for newsletter "Interesting Reads" section to eliminate AI hallucinations in citations.

Problem: jekyll-scholar requires Jekyll 4.x but GitHub Pages uses 3.10.0 (incompatible).

Solution: BibTeX → YAML → Liquid template pipeline.

Implementation

• scripts/bibtex_to_yaml.py - Parses BibTeX, formats authors (abbreviates >3), generates YAML
• _includes/bibliography-entry.html - Liquid template matching newsletter style (bold authors, italicized titles/venues)
• _data/interesting_reads.bib - Sample bibliography exported from Zotero
• _template/newsletter-template.md - Auto-renders entries filtered by read_status=Unread

Workflow

Zotero collection → Export BibTeX → scripts/bibtex_to_yaml.py → Jekyll renders

Output Example

BibTeX:

@inproceedings{obrien2025llm,
  title = {How Scientists Use Large Language Models to Program},
  author = {O'Brien, G.},
  booktitle = {CHI '25: Proceedings of the CHI Conference on Human Factors in Computing Systems},
  pages = {1--16},
  year = {2025},
  doi = {10.1145/3706598.3713668},
  read_status = {Unread}
}

Renders as:

G. O'Brien, "How Scientists Use Large Language Models to Program," CHI '25: Proceedings of the CHI Conference on Human Factors in Computing Systems, pp. 1–16, 2025. Read the article.

Documentation

• _data/README_BIBLIOGRAPHY.md - Complete workflow
• _data/BIBLIOGRAPHY_EXAMPLES.md - Examples and troubleshooting
• IMPLEMENTATION_SUMMARY.md - Technical details

Fixes #[issue-number]

Checklist:

• I have previewed changes locally or with CircleCI (runs when PR is created)
• I have completed any content reviews, such as getting input from relevant working groups. If no, please note this and wait to post the PR to the #website channel until the content has been settled.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Use Zotero and Jekyll-scholar to automate "interesting reads" section (NO AI)</issue_title>
<issue_description>Due to the previous embarrassing snafu that AI was hallucinating URLs of articles we asked it to format, I want to go the pure code route. The plan is as follows:

1. Throughout the month, collect interesting media, be it official peer reviewed publications, blog posts, or videos, in zotero with a dedicated bibtex bibliography. This zotero collection could even be shared with the community as a group.
2. In zotero, use a special tag (such as the read status plugin https://github.com/Dominic-DallOsto/zotero-reading-list ) to categorise what has already been published and what is waiting for the next edition
3. When the time to publish comes, export the full bibliography bibtex file to the repo/update the bibtex file in the repo (e.g. in _data)
4. In the website repo, use the jekyll-scholar plugin syntax https://github.com/inukshuk/jekyll-scholar to query the bibliography for zotero entries that have NOT been read by the read status plugin

# Interesting Reads

{% bibliography --query @*[read_status=Unread] %}

4. Customise the CSL so that entries in the above bibliography come out in the way that we have been shaping them for a newsletter, eg:

- **G. O’Brien**, _"How Scientists Use Large Language Models to Program,"_ *CHI ’25: Proceedings of the CHI Conference on Human Factors in Computing Systems*, pp. 1–16, 2025. [Read the article.](https://doi.org/10.1145/3706598.3713668)
</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Use Zotero and Jekyll-scholar to automate "interesting reads" section (NO AI) #51

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.