
TL;DR
Justin Poehnelt spent seven years at Google building open-source developer tools. His CLI went viral, hit #1 on Hacker News, and got him fired two days before Google announced their own version.
Two months ago, Justin Poehnelt was fired by Google. His offense: creating a CLI tool for Google Workspace that went viral, hit #1 on Hacker News, accumulated thousands of GitHub stars, and attracted many thousands of users in just a couple of days.
The timing made it worse. Two days before his termination, Google announced at Cloud Next 2026 that an official Workspace CLI was in development.
Poehnelt spent nearly seven years on Google's Workspace Developer Relations team. His job involved building open-source layers and abstractions over Google APIs - exactly the kind of work the team was designed to do. The CLI tool he created, gws, provides unified command-line access to Google Drive, Gmail, Calendar, and other Workspace APIs.
When the tool went viral, it caught leadership by surprise. Directors and leaders asked what they could learn from the project. Then legal started grilling him about why Google's logo and brand colors appeared on a Google Workspace GitHub code repository.
According to Poehnelt, the core issue was fear of disruption - not about his CLI specifically, but about what AI agents meant for Workspace as a product. The CLI made Workspace APIs accessible in a way that AI coding assistants and automation tools could easily consume.
The Hacker News thread has 340+ comments and surfaces several angles on the story.
The process question dominates. Multiple former Googlers confirm that Google has strict OSS release processes. The debate is whether Poehnelt followed them. He claims the process is "not clearly documented and always changing" and that he had approval through the internal launch system (Ariane/Launcher2) with the engineering bit flipped by his manager. Others point to Google's public OSS release documentation as evidence the rules were clear.
The branding issue is murky. The GitHub organization googleworkspace displays Google's logo on all its repositories - that's an org-level setting, not something Poehnelt added. The README includes the standard "This is not an officially supported Google product" disclaimer. But releasing a viral product with official-looking branding without the full corporate launch review is risky at any large company.
20% time is invoked nostalgically. Several commenters see this as evidence that Google's famous 20% time culture is dead. "Google has gone from encouraging 20% time to firing people for doing it," writes one commenter. Others push back: 20% time never meant bypassing launch approvals, and this appears to be less about side projects and more about proper channels.
The AI disruption angle resonates. The CLI made Google Workspace APIs trivially accessible to AI agents. One commenter notes: "Your tool is something that made Workspace so much more useful to me personally... Getting fired for making a product more useful to customers is quite ironic." Another adds that paired with a Claude skill, it saved significant time creating meeting notes - exactly the kind of AI-native workflow that Workspace apparently wasn't ready to officially support.
Corporate politics gets blamed. "Good ideas are now risky because it steps on the toes of someone's fiefdom," writes one commenter. Another: "They've been GE'd." The general sentiment is that something broke in how Google handles internal innovation.
Read the full discussion at Hacker News.
Get the weekly deep dive
Tutorials on Claude Code, AI agents, and dev tools - delivered free every week.
From the archive
Jun 24, 2026 • 7 min read
Jun 23, 2026 • 7 min read
Jun 23, 2026 • 8 min read
Jun 23, 2026 • 8 min read
This situation illustrates a recurring tension in big tech: the gap between what developer relations teams are supposed to do (build tools that make platforms accessible) and what product teams want to control (the timing, branding, and narrative around new capabilities).
It also highlights how AI is changing developer tools. A CLI that exposes APIs cleanly isn't just a convenience anymore - it's infrastructure for AI agents. When every developer has access to coding assistants that can call arbitrary APIs, making those APIs easily callable becomes a strategic decision.
For developers working at large companies, the lessons are practical:
Document your approvals. If you have sign-off, make sure it's on record and that you understand exactly what scope it covers.
Understand branding implications. Using company GitHub orgs, logos, or anything that could make your project look official creates liability. Even "not officially supported" disclaimers may not be enough if the visual presentation suggests otherwise.
Consider timing. A project that goes viral right before your company announces a competing official version creates an awkward situation for everyone - especially if your project is better received.
Recognize disruption risk. If your side project enables use cases that threaten existing business models (like AI agents automating away SaaS seats), expect friction from stakeholders who see the threat before they see the opportunity.
The Google Workspace CLI (gws) is still available at github.com/googleworkspace/cli. It provides command-line access to Workspace APIs in a format that works well with AI coding tools. Whether Google eventually releases their own version or claims this one remains unclear.
For now, it serves as a case study in what happens when developer tools become too useful too fast.
Read next
The Transformer co-creator leaves Google DeepMind for OpenAI just two years after Google paid $2.7 billion to bring him back from Character.AI.
5 min readEpic Games open-sourced Lore, a centralized version control system designed for binary-heavy game projects. It uses Merkle trees, on-demand file hydration, and native chunked storage to handle terabyte-scale repos that Git struggles with.
7 min readA developer used OpenAI Codex to build a fully open-source WYSIWYG editor for TikZ figures. The technical approach and reception on Hacker News offer a useful case study in what agent-built software looks like when shipped.
7 min readTechnical content at the intersection of AI and development. Building with AI agents, Claude Code, and modern dev tools - then showing you exactly how it works.
The TypeScript toolkit for building AI apps. Unified API across OpenAI, Anthropic, Google. Streaming, tool calling, stru...
View ToolGoogle's open-source coding CLI. Free tier with Gemini 2.5 Pro. Supports tool use, file editing, shell commands. 1M toke...
View ToolUnified API for 200+ models. One API key, one billing dashboard. OpenAI, Anthropic, Google, Meta, Mistral, and more. Aut...
View ToolGoogle's frontier model family. Gemini 2.5 Pro has 1M token context and top-tier coding benchmarks. Gemini 3 Pro pushes...
View Tool
A developer used OpenAI Codex to build a fully open-source WYSIWYG editor for TikZ figures. The technical approach and r...

Baidu releases Unlimited OCR, an open-source vision-language model that parses 100+ page documents in a single pass with...

Switzerland's fully open foundation model promises transparent training data and EU compliance. The HN crowd has questio...

A YC W25 startup open-sources CADAM, a browser-based tool that converts natural language to parametric OpenSCAD models....

The Transformer co-creator leaves Google DeepMind for OpenAI just two years after Google paid $2.7 billion to bring him...

Epic Games open-sourced Lore, a centralized version control system designed for binary-heavy game projects. It uses Merk...

New tutorials, open-source projects, and deep dives on coding agents - delivered weekly.