Changelog

Follow the evolution of ChatEngram, from the first public release onward.

Version v1.2.7

Smarter Screenshots, Better Popup UI, and Stability Improvements

2026-01-23Status:Rolling out

This release improves screenshot quality, reduces noisy errors, and makes the popup faster and easier to use — without changing how your capture and search workflows work.

✨ Improvements

📸 Perplexity screenshots now include Sources

When you generate a smart screenshot for a Perplexity conversation, the screenshot can include the sources list so your references are preserved when sharing.

🧩 Popup UI refinements

The popup has been polished for quicker access to your dashboard, recent items, and key actions.

🛠 Fixes & Reliability

  • Reduced console noise and improved long-session stability
  • Improved reliability of plan/status syncing across extension pages
  • History backfill prompts are less repetitive

Version v1.2.0

ChatGPT 5.2 Support & Database Stability

2026-01-02Status:Released

This update adds full support for ChatGPT 5.2 and resolves database initialization issues that could prevent proper conversation syncing.

✨ New Features

🚀 ChatGPT 5.2 Support

ChatEngram now fully supports the latest ChatGPT 5.2 release. All conversation capture, streaming, and history backfill features work seamlessly with the new version.

🔧 Bug Fixes
  • Fixed database initialization issues that could cause empty or corrupted storage
  • Improved database version migration from older installations
  • Enhanced error handling for database operations
⚡ Performance Improvements
  • Removed debug logging to improve performance
  • Optimized content synchronization for faster message capture
  • Reduced memory usage during streaming operations
  • Improved extension startup time

Version v1.1.1

Hotfix: ChatGPT page load regression

2025-12-31Status:Released

This hotfix resolves a regression where visiting chatgpt.com with the ChatEngram extension enabled could show a "Content failed to load" error.

The extension's network interceptor no longer treats ChatGPT's own backend API endpoints as third-party tracking calls, so the site can load normally while still suppressing real analytics/advertising requests.

What changed

  • Fixed an issue where wrapped window.fetch and XMLHttpRequest could return empty 204 responses for critical ChatGPT API calls.
  • Tightened the tracking-domain filter so only real analytics and ad domains are short-circuited; ChatGPT and Chat.OpenAI domains now pass through untouched.
  • History backfill and streaming capture continue to work as before on supported platforms.

Version v1.1.0

New Logo, AI Summary & Extension Improvements

2025-12-29Status:Released

This release focuses on making the browser extension more useful out of the box, with a refreshed logo, built-in AI summaries, and better control over which models you use.

New in the browser extension

New extension logo

The ChatEngram extension now ships with a refreshed logo that looks sharper in the browser toolbar, popup, and Chrome Web Store. It's easier to recognize at a glance and better aligned with the overall product brand.

AI Summary inside the dashboard

You can now generate an AI summary for any conversation directly from the dashboard detail view. Summaries are cached per conversation so revisiting them doesn't incur extra API cost.

Multiple styles are supported: quick summary, detailed summary, key points, action items, and fully custom prompts for advanced users.

Configurable providers and low-cost defaults

A dedicated AI Summary Settings panel lets you choose between OpenAI, Anthropic, Google Gemini, and DeepSeek, and manage your own API keys.

Default models are optimized for speed and cost (gpt-4o-mini, claude-3-haiku-20240307, gemini-1.5-flash, deepseek-chat), and the Model field is fully editable so you can plug in any other supported model name when needed.

Clearer cost hints

The extension now shows approximate per-summary costs for each provider, along with a reminder to verify pricing on the official dashboards, helping you keep API usage under control.

Version v1.0.2

Multi-language Support & Citation Fixes

2025-12-22Status:Released

This update brings comprehensive multi-language support and fixes citation display issues for Perplexity conversations.

✨ New Features

🌍 Multi-language Support (20 Languages)

ChatEngram now speaks your language! The entire interface automatically adapts to your browser settings.

Supported languages: English, Chinese (Simplified & Traditional), Japanese, Korean, Spanish, Portuguese, German, French, Italian, Russian, Swedish, Finnish, Danish, Polish, Dutch, Norwegian, Turkish, Hebrew, and Thai.

🔧 Bug Fixes
  • Fixed missing citation sources in Perplexity conversation details
  • Fixed citation display in smart screenshots for Perplexity conversations
  • Improved version number display in dashboard (now reads from manifest automatically)

Version v1.0.0

Initial Release

2025-12-06Status:Public Launch (Chrome Web Store)

The first official version of ChatEngram is now live. This release marks the foundation of a fully local-first AI memory system — built for users who rely on ChatGPT, DeepSeek, Claude, Gemini, and other AI platforms every day.

This initial version includes all core features that define the product vision.

✨ Core Features

1. Automatic AI Conversation Capture

Automatically saves your conversations across supported platforms, including ChatGPT, DeepSeek, Claude, Gemini, and Perplexity. Runs quietly in the background — no manual export needed.

2. Unified Global Search

A single search box for all your AI conversations — across all platforms. Instant indexing. Local-only. Privacy guaranteed.

3. Spotlight Command Bar (Cmd/Ctrl + K)

Navigate your entire AI memory instantly. Jump to any conversation or message in milliseconds with a universal command palette.

4. Knowledge Sync

Effortlessly sync all chats to your knowledge tools:

  • Obsidian Markdown vaults
  • Notion
  • Google Drive

Each conversation is exported as clean, structured Markdown — with timestamps and metadata.

5. Multi-chat Export

Select multiple conversations and export them together as a JSON backup file.

Perfect for research, writing, archiving, or processing with your own tools.

6. History Backfill

Automatically scrolls through older conversations to capture your complete history — even from before installing the extension.

7. Local-first Storage

Your AI memory stays on your device:

  • No servers
  • No cloud uploads
  • No tracking
  • No account required

This is privacy by design.

8. Adapter Lab (Advanced)

For technical users:

  • Custom DOM selectors
  • Custom parsing rules
  • Extend support to new or niche AI platforms
  • Self-healing when UIs update

This makes ChatEngram one of the most extensible AI memory tools on the web.