Syntax - Tasty Web Development Treats
Syntax - Tasty Web Development Treats

1018: Google fires Workspace CLI Creator

July 6, 2026

AI Summary

5 min read

Google fired the creator of its own Workspace CLI, OpenAI previewed GPT‑5.6 in three tiers, and Astro 7 rewrote its component compiler in Rust. The episode also covered Cursor’s new iOS app, Coinbase’s model-routing cost cuts, a privacy-focused Android keyboard, and a handful of smaller releases.

Google Fires the Creator of the Workspace CLI

Justin Ponel, a Google employee, built and released a CLI for Google Workspace that let users interact with Calendar, Gmail, and other products from the terminal. The tool landed on what appeared to be an official Google GitHub organization, was promoted by Google’s own DevRel, and went viral — Wes Bos said his own post about it got over a million views. “Everybody was like, finally, this is exactly what I want,” Wes said. “I want a CLI that I can give to my agents to be able to interact with all of my Google products.”

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

What you'll learn

  • 1 (01:47) **Google Fires Workspace CLI Creator** - Justin Ponel, creator of the Google Workspace CLI, was fired after releasing the tool, which went viral and was widely used.
  • 2 (12:28) **GPT-5.6 Previewed** - OpenAI previewed three new models in the GPT-5.6 series: Sol, Terra, and Luna.
  • 3 (20:02) **Open-Weight Model GLM-5.2 Released** - A large open-weight model from Zhipu AI, on par with frontier models like Opus 4.6 and GPT-5.5.
  • 4 (23:22) **Astro 7 Rewritten in Rust** - The Astro web framework released version 7, with its component compiler rewritten in Rust for major performance gains.
  • 5 (32:45) **Cursor iOS App Released** - Cursor launched an iOS app, allowing remote control of cloud agents and local sessions from a phone.
  • 6 (40:35) **Coinbase Cuts AI Spend with Model Routing** - Coinbase reduced its AI spend nearly in half by routing simpler tasks to cheaper open-weight models via an LM gateway.
  • 7 (48:00) **GitHub Copilot Token Efficiency** - Data shows GitHub Copilot uses significantly fewer tokens (up to 50-60% less) than native model harnesses (e.g., Claude Code, GPT) for similar outputs.

+ Full timestamped outline available in the app

Guests on this episode

Show Notes

Google fires the engineer behind its Workspace CLI tool, OpenAI previews GPT-5.6 with three new model tiers, and Astro 7 lands with a full Rust rewrite. Plus: Coinbase cuts token costs with smarter routing, and more in this week’s Syntax Live

Show Notes

Syntax - Tasty Web Development Treats