Sync Gemini chat logs directly to Google Sheets & Drive in real-time.
Built on Manifest V3 with zero infrastructure costs — perfect for Indie Hackers & Engineers.
A clean, production-ready codebase powered by Chrome Manifest V3 and Google Apps Script (GAS). Automatically formats Gemini responses into clean Markdown and syncs structured data into Google Sheets & Drive in real-time.
A solid base to build and launch custom AI / Gemini tools rapidly.
Clean design pattern integrating Chrome Manifest V3 with GAS.
Automate manual copy-pasting by syncing AI output directly to Drive.
Automation & System Design Insights
The final installment on gaining the self-reliance to remain unfazed even when 'the tool that worked yesterday suddenly stops'! I'll teach you the independent design of parser.js (assuming it will break) and the steps to give precise fix instructions to AI when the UI changes. Here's a summary of the thinking process to safely maintain and operate tools without completely leaving everything to AI.
Taking on the challenge of fully automating the Gemini chat log saver (Part 3). I detail the process of avoiding the risks of periodic execution every 1.5 seconds, and building a smart DOM monitoring process (MutationObserver) that doesn't burden the other party's server. I'll share the thinking process of having the human point out risks and making the AI come up with safe alternatives.
Solved the 'long texts and code are hard to read' issue in spreadsheet saves! I expanded GAS to build a system that auto-generates Markdown (.md) files to Google Drive. I'll explain the process of fixing a line-break bug with AI, and the ingenuity behind a 'robust design' resistant to UI changes.
Avoid the trap of 'I made the AI write the code, but it doesn't match the actual screen and won't work'! I'm revealing the development process behind a tool that automatically saves Gemini chat logs. In Part 1, I'll show you how to effectively control AI through understanding system structure and giving precise instructions using DevTools.
e-Shikumi-labo