✨ GitHub Copilot Chat Exporter ✨

Export GitHub Copilot shared conversations to clean Markdown with inline attachments. No API key required!

A Python-based toolkit to extract and archive GitHub Copilot chat share pages with full authentication support, file attachment capture, and inline file linking.

Features

• ✅ Export Copilot shared conversations to Markdown
• ✅ File attachment capture - Extracts CSV, TXT, JSON, YAML, XML, MD, and code files
• ✅ Inline attachment links - Files appear where they're referenced in the conversation
• ✅ Authenticated access via manual login (saves reusable session state)
• ✅ Headless automation with Playwright
• ✅ Fallback static scraper for public pages
• ✅ JSON API extraction with DOM fallback
• ✅ Debugging artifacts (page.html, network logs)
• ✅ Clean, simple folder structure (attachments only)
• ✅ Works on macOS, Linux, and Windows

Installation

macOS/Linux (Homebrew) - Recommended

brew tap pandaxbacon/tap
brew install github-copilot-chat-exporter

That's it! The copilot-exporter command is now available.

All Platforms (Manual Setup)

Prerequisites

• Python 3.9 or higher
• pip (Python package installer)

Setup

1. Create a virtual environment:

python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

2. Install dependencies:

pip install -r requirements.txt
python -m playwright install chromium

Quick Start

Homebrew Users

1. Authenticate (One-time setup):

copilot-exporter --mode login --url https://github.com/copilot/share/YOUR-SHARE-ID

• A browser window will open
• Sign in to GitHub
• Press Enter in the terminal to save authentication

2. Export Conversations:

# Markdown only (text and code blocks)
copilot-exporter --mode run --url https://github.com/copilot/share/YOUR-SHARE-ID

# With file attachments (captures CSV, TXT, JSON, YAML, code files, etc.)
copilot-exporter --mode run --url https://github.com/copilot/share/YOUR-SHARE-ID --with-assets

Manual Setup Users

1. Authenticate (One-time setup):

python scraper_playwright.py --mode login --url https://github.com/copilot/share/YOUR-SHARE-ID

• A browser window will open
• Sign in to GitHub
• Press Enter in the terminal to save storage_state.json

2. Export Conversations:

python scraper_playwright.py --mode run --url https://github.com/copilot/share/YOUR-SHARE-ID

# With CSV attachments (captures uploaded and generated files)
python scraper_playwright.py --mode run --url https://github.com/copilot/share/YOUR-SHARE-ID --with-assets

Output Files

• chat-export.md - Clean Markdown format with inline attachment links
• page.html - Full page HTML for debugging
• storage_state.json - Saved authentication state (reusable)
• Asset folder (when using --with-assets):
  • output/<conversation>/attachments/ - CSV files and user uploads with inline links

Usage

Playwright Scraper (Recommended)

The Playwright-based scraper supports authenticated access and renders JavaScript-heavy pages.

Authentication Mode

python scraper_playwright.py --mode login --url <SHARE_URL>

Opens a headed browser for manual GitHub login, then saves authentication to storage_state.json.

Export Mode

python scraper_playwright.py --mode run --url <SHARE_URL> [--with-assets]

Runs headless using saved authentication. Exports to Markdown with optional attachment capture.

Options:

Flag	Description	Default
--mode	login or run	run
--url	GitHub Copilot share URL	Sample URL
--with-assets	Capture CSV files and attachments	False

Requests Scraper (Public pages only)

For publicly accessible share pages (no authentication required):

python scraper_requests.py --url <SHARE_URL>

Note: This will exit with a warning if the page requires login. Use the Playwright scraper for authenticated pages.

Examples

Basic Export

# Assuming storage_state.json exists from login
import asyncio
from pathlib import Path
from scraper_playwright import run_export

asyncio.run(run_export(
    url="https://github.com/copilot/share/YOUR-SHARE-ID",
    with_assets=False  # Set to True to capture CSV files
))

Batch Export Multiple URLs

import asyncio
from scraper_playwright import run_export

urls = [
    "https://github.com/copilot/share/ID-1",
    "https://github.com/copilot/share/ID-2",
    "https://github.com/copilot/share/ID-3",
]

async def batch_export():
    for url in urls:
        await run_export(url, with_assets=True)
        print(f"✓ Exported {url}")

asyncio.run(batch_export())

Output Format

Markdown Structure

# Chat Export

## User
we should not pursue a perfect cut
instead we may run multiple runs … and then add another workflow at the end to judge and vote which cut is the best

## Copilot
That's exactly the right instinct:
- There isn't one optimal chunking for all tasks.
- You've already seen two "valid but different" interpretations of the same document.
...

Attachments & Folder Structure (--with-assets)

When --with-assets is set, exports are organized under output/<conversation>/:

output/
  conversation-title/
    chat-export.md           # Markdown with inline attachment links
    page.html                # Debug artifact
    attachments/
      aia-glossary-001.csv            # User-uploaded CSV
      aia-glossary-cleaned-001.csv    # Copilot-generated CSV
      image-001.png                   # User-uploaded image

Inline Attachment Links:

Attachments appear directly where they're referenced in the conversation:

## User
📎 **Attachment:** [aia_glossary.csv](attachments/aia-glossary-001.csv)

here you are file

## Assistant  
Here's the cleaned file I produced:

📎 **Attachment:** [aia_glossary_cleaned.csv](attachments/aia-glossary-cleaned-001.csv)

See examples/sample-export/ for a complete real-world example.

Troubleshooting

"storage_state.json not found"

Solution: Run the login mode first:

python scraper_playwright.py --mode login --url <SHARE_URL>

Empty or missing messages

Possible causes:

1. Selectors changed: GitHub may have updated the page structure

   • Check page.html to inspect the rendered DOM
   • Update MESSAGE_SELECTORS in scraper_playwright.py

2. Network timing: Page didn't fully load

   • Increase wait timeout in run_export() function
   • Run in headed mode (headless=False) to visually confirm rendering

"Page redirected to login"

Cause: Authentication expired or not captured

Solution: Re-run login mode to refresh storage_state.json

Platform Support

Platform	Status
✅ macOS	Fully supported
✅ Linux	Fully supported
✅ Windows	Fully supported

Development

Running Tests

# Install dev dependencies
pip install -r requirements-dev.txt

# Run tests
pytest

# Run tests with coverage
pytest --cov=. --cov-report=html

# View coverage report
open htmlcov/index.html

Code Quality

# Format code
black scraper_playwright.py scraper_requests.py

# Lint
flake8 scraper_playwright.py scraper_requests.py

# Type check
mypy scraper_playwright.py scraper_requests.py

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

See CONTRIBUTING.md for guidelines.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Credits

• Built with Playwright for browser automation
• Inspired by the need to archive AI-assisted development conversations

Support

• 📖 Documentation: See developer-notes/ for R&D details
• 🐛 Issues: Report on GitHub Issues
• ⭐ Like this project? Give it a star!

---

Made with ❤️ for developers who want to own their AI conversation history