t3.unchat

Convert t3.chat JSON data exports into structured markdown and styled HTML files for offline browsing and archival.

Overview

t3.chat provides a JSON export of all your conversations. This tool converts that export into a browsable collection of markdown and HTML files organized by date. Each conversation thread becomes a pair of files (.md and .html) in a YYYY/mm/ directory structure, with a top-level index.html for navigation.

Disclaimer

This program was vibe-coded by Claude Opus 4.6. As such, the author can't be held responsible for incorrect results. Please verify any and all output.

Features

• Parses the threads-export-YYYY-mm-ddT1HH_MM_SS.sssZ.json export and maps messages to their corresponding threads
• Generates one markdown file per thread with user/assistant message pairs
• Produces styled HTML with syntax-highlighted code blocks (via Pygments)
• Creates a top-level index.html with all threads grouped by month in reverse chronological order, including within the same day
• Pinned threads are marked with a pin icon in both the HTML page title and the index
• Each markdown file includes a Metadata section with the full original thread JSON and a Computed Date Times section with human-readable timestamps
• Handles duplicate thread titles with numeric suffixes (e.g., Title_2.md)
• Case-insensitive duplicate detection for macOS/Windows filesystem compatibility
• Filenames are sanitized to remove problematic characters, with spaces converted to underscores and a maximum length of 80 characters
• Configurable timezone for computed datetimes (defaults to US/Eastern)

Attachment Limitations

The t3.chat JSON export includes attachment IDs referenced by messages, but there is no way to map these attachment IDs to actual files. The export does not contain attachment content, filenames, URLs, or any other recoverable data about the attachments. Messages that referenced attachments will display placeholder text like [Attachment: j456jsa24jabccv0thtfzexyz123d27s] in the output.

Requirements

• Python 3.12+
• uv for dependency management

Dependencies

• markdown -- markdown to HTML conversion with fenced code block, table, and code highlighting extensions
• Pygments -- syntax highlighting for code blocks in the HTML output

Installation

git clone https://github.com/jftuga/t3.unchat.git
cd t3.unchat
uv sync

Usage

Step 1: Export markdown files

uv run python t3export.py threads-export-YYYY-mm-ddT1HH_MM_SS.sssZ.json -o output

Options:

Flag	Description
-o, --output	Root output directory (default: output)
--utc	Display computed datetimes in UTC instead of US/Eastern

Step 2: Generate HTML files and index

uv run python t3html.py output -s threads-export-YYYY-mm-ddT1HH_MM_SS.sssZ.json

Options:

Flag	Description
-s, --source	Source export filename to display in the index header

This converts every .md file to a companion .html file with embedded CSS and syntax-highlighted code blocks, then generates a top-level index.html.

Browsing the output

Open output/index.html in a web browser. Threads are grouped by month in reverse chronological order. Each entry links to the styled HTML version of the conversation.

Output structure

output/
  index.html
  2025/
    03/
      20250308--Thread_Title.md
      20250308--Thread_Title.html
      20250309--Another_Thread.md
      20250309--Another_Thread.html
    04/
      ...
  2026/
    ...

Filename format

Each file is named YYYYmmdd--Sanitized_Title.md where:

• YYYYmmdd is the thread creation date
• -- is a literal separator
• Sanitized_Title has unsafe characters removed, spaces replaced with underscores, and is truncated so the full stem stays within 80 characters

Markdown file structure

Each markdown file contains:

1. Thread title as an h1 heading
2. Messages alternating between ## User and ## Assistant sections, separated by horizontal rules
3. Metadata section with the complete original thread JSON
4. Computed Date Times section with deduplicated timestamps converted to human-readable format in the configured timezone

Project structure

Files are listed in recommended study order (foundation-first, complexity-last):

1. t3_thread.py -- T3Thread dataclass representing a conversation thread. Handles parsing from JSON, filesystem-safe filename generation, metadata serialization, and computed datetime conversion.
2. message.py -- Message dataclass representing a single user or assistant message. Handles parsing from JSON and rendering to markdown with attachment placeholders.
3. exporter.py -- Exporter class that orchestrates the full export. Loads the JSON, groups messages by thread, resolves duplicate filenames, and writes the markdown files to the date-based directory structure.
4. t3export.py -- CLI entry point for the markdown export step. Parses arguments and delegates to Exporter.
5. t3html.py -- CLI entry point for the HTML conversion step. Converts markdown files to styled HTML with syntax highlighting, adds pin icons to pinned threads, and builds the index.html navigation page.

Personal Project Disclosure

This program is my own original idea, conceived and developed entirely:

• On my own personal time, outside of work hours
• For my own personal benefit and use
• On my personally owned equipment
• Without using any employer resources, proprietary information, or trade secrets
• Without any connection to my employer's business, products, or services
• Independent of any duties or responsibilities of my employment

This project does not relate to my employer's actual or demonstrably anticipated research, development, or business activities. No confidential or proprietary information from any employer was used in its creation.