Google fired the engineer behind its 20k-star Workspace CLI
He shipped a Rust CLI for Google Workspace built entirely with AI agents. It hit #1 on Hacker News and 20,000 stars. Then Google fired him over it.
Justin Poehnelt, a senior devrel engineer, published an open-source Google Workspace CLI under the company's own GitHub org, wearing Google's logo and brand colors. It hit 20k+ stars and #1 on HN. He says legal grilled him over the branding, and the deeper fear was what agents meant for Workspace.
A former Google Chrome engineer in the thread complicated the cheering: publishing under official Google orgs was routine, but shipping something that looks like an official product without sign-off is the kind of lapse that gets you walked out.
A side project that hit #1, then ended a career

Justin Poehnelt, a Senior Developer Relations Engineer at Google, posted on X that he was fired over a side project: an open-source Google Workspace CLI he wrote in Rust, built entirely via AI agents. His own profile records the outcome plainly — the tool "hit 20k+ stars and #1 on HN." It also lived under the googleworkspace GitHub org and wore Google's logo and brand colors.
By his account the trouble started with legal: he was questioned about why Google's logo and brand were on those repositories. But he frames the real driver as something larger than trademark — "the fear wasn't specific to my CLI, it was a broader fear in what agents meant for Workspace." An autonomous CLI that lets agents drive Gmail, Docs and Drive is, after all, a direct threat to how Workspace is sold.
Poehnelt declined to elaborate, writing only that the episode "speaks to the experience of working in big tech and the disruption caused by AI." That silence became its own subplot in the thread.
The Xooglers split on whether this was insane or deserved
The sympathy was loud at first, then a former insider muddied it. A developer who worked in Google's Chrome org pushed against the idea that publishing was the crime:
"when I worked at Google under Chrome it was very common for individuals and teams to publish projects to open source repositories under Google-managed Github orgs… Firing seems an extreme consequence" — a former Chrome engineer, 2015–2021
Another ex-Googler was blunter about the optics:
"This is exactly the kind of shit-for-brains action I'd expect from Google executives." — a former Google employee
But the counter-read was just as concrete, and it came down to brand impersonation. One commenter admitted the project fooled him:
"The lack of judgement involved in personally releasing something that could be confused for an official release (I was confused) by your employer is someone who has huge wildcard risk" — a commenter who mistook the CLI for an official Google release
The README itself says "This is not an officially supported Google product" — which several readers found contradictory next to the Google branding and a Google employee's name. A developer at a much smaller shop noted that even there, creating a public repo requires review: how did an unapproved public repo under Google's org happen at all?
The detail everyone wanted Poehnelt to answer
The sharpest pushback wasn't about policy — it was about timing, and it was aimed at Poehnelt directly for posting the story and then going quiet:
"Why the hell is someone on our team releasing a vibe-coded CLI that's branded to look like an official API, when we're 2 weeks from announcing the actual CLI?" — khazhoux, pressing the unanswered questions
If an official Workspace CLI was already in the pipeline, an identically-scoped unofficial one from inside the same org reads very differently than "Google fires builder." Others who'd been in similar spots described the safe path: one engineer reverse-engineered a private API into a useful CLI, sat on it, and only got it open-sourced — plus an innovation award — after legal fast-tracked the review. The lesson he drew: pushing it to a public repo without that review is "suicidal."
The through-line isn't whether Google was petty or correct — it's that a single screenshot-thread on X can't settle it, and the engineer who has the missing facts chose not to share them. What the thread does establish: at a company once famous for 20% time, a hit side project is now a liability worth a termination, and even the Xooglers who'd defend the culture can't agree on whether the rules changed or the developer broke them.
Get breakdowns like this every day
The top Hacker News stories, unpacked — by email, Telegram, or RSS.
Anthropic's new terms can ask for your ID or your facial geometry
Anthropic's updated privacy policy spells out what 'verify your age or identity' can mean: an image of your government ID, a photo or video of your face, and facial-geometry templates that may count as biometric data. The thread's first real correction: this isn't news - identity checks have been quietly catching users for months.