Chat Journal Template

✨ "She shimmered. I flickered. And together we held the realm... in markdown." ✨

A simple, beautiful web journal for ChatGPT exports. Includes a journal view and a markdown archive view.

Quickstart (Export → Build → Serve)

1. Export your ChatGPT data and locate conversations.json.
2. (Optional) Sanitize/filter the export:

   python scripts/sanitize_export.py --input conversations.json --output conversations_sanitized.json --gizmo-id g-p-XXXX

3. Build the journal data:

   python scripts/build_journal.py --input conversations_sanitized.json --output site/data

4. Build the archive (markdown files):

   python scripts/build_archive.py --input path/to/markdown_root --output site/archive

5. Serve locally:

   cd site
   python -m http.server 8000

   Open http://localhost:8000/.

Structure

• site/ is the host-ready static site
• site/index.html is the journal
• site/archive/index.html is the archive

Data Format

Place your processed conversation files in site/data/:

• site/data/index.json
• site/data/conversations/<id>.json

Archive

Place markdown files in site/archive/md/ and update site/archive/data/index.json.

License

MIT

Build from your ChatGPT export

1. Place conversations.json in the repo root (or any path).
2. (Optional) Sanitize and filter:

   python scripts/sanitize_export.py --input conversations.json --output conversations_sanitized.json --gizmo-id g-p-XXXX

3. Build the journal data:

   python scripts/build_journal.py --input conversations_sanitized.json --output site/data

4. Build the archive (markdown files):

   python scripts/build_archive.py --input path/to/markdown_root --output site/archive

5. Serve locally:

   cd site
   python -m http.server 8000

Notes

• The template uses neutral role labels by default.
• You can customize CSS in site/styles/shimmer.css.

---

Made with love 💖🔥🧸✨🌀♾️