Grok BotGrok Bot has only been public since August 2026, but the interesting part is already moving beyond demos.
People are not only asking Bots questions. They are giving them persistent jobs, connecting them to tools, having specialized Bots hand work to one another, and publishing the systems that result.
Some examples below come directly from xAI. Others are open-source projects or community builds. Those distinctions matter: an official case tells us what the product is designed to support; a community post tells us what someone says they built, not necessarily what has been independently verified.
1. A six-Bot mobile game studio (opens in a new tab)
One of xAI’s most concrete examples is Rank’em, a mobile game operated with a team of six Grok Bots. Different Bots handle different parts of the work around the app, including analytics and user acquisition. The author reported that the game had crossed 1,000 iOS and Android downloads during the week described in the guide.
The useful lesson is not “AI built a game.” It is the division of labor. One persistent agent is not being asked to remember every job in the company.
2. An engineering team managing other coding agents (opens in a new tab)
An xAI engineering guide describes Grok Bots supervising cloud coding agents, checking work and keeping engineering jobs moving while the human is away. The post also reports significant internal development output, including an early Grok Bot iOS build completed using Grok Bot.
This suggests a different use for persistent agents: managing agent work, not necessarily writing every line themselves.
3. A six-seat marketing department (opens in a new tab)
Another official guide describes six marketing roles: Market Researcher, Product Marketer, Performance Marketer, Website Ops, Marketing Analyst and Project Manager.
Research can flow into positioning in Google Docs, approved copy into Google Ads structures, page briefs into code and GitHub PRs, and campaign data back into analysis. The human still owns the judgment layer and live-account approvals.
4. A GTM team that works overnight (opens in a new tab)
xAI’s GTM guide includes a Chief of Staff, prospecting Bot, customer experts, forecasting, slides and sales coaching. The prospecting setup can research accounts overnight and leave material ready for review the next morning.
5. A PM with an agent team (opens in a new tab)
xAI product manager Kevin Niparko describes using specialized Bots for engineering management, engineering, analytics, product work and recruiting. One particularly interesting idea is an attention list generated from the work actually happening across Slack, email, meetings and calendar activity.
6. Projects Manager (opens in a new tab)
The official Marketplace includes Projects Manager, which uses Notion projects and tasks as an operating layer for specialist Bots. The manager coordinates; the specialists execute.
7. SEO & AEO Desk (opens in a new tab)
The Marketplace’s SEO & AEO Desk converts keyword or Search Console input into content ideas and writer-ready briefs.
8. Stalk Bot (opens in a new tab)
Stalk Bot takes competitive research unusually far: its Marketplace description says it can sign up for competitor newsletters and products, walk onboarding, and monitor pricing, changelogs, hiring and X.
9. GrokBotDev (opens in a new tab)
The open-source GrokBotDev project turns Grok Bot discovery itself into an agent-run operation. Scout agents look for public examples, a curator decides what belongs, and a builder maintains the site through reviewable pull requests.
10. UseGrokBot (opens in a new tab)
UseGrokBot is another open-source attempt to organize the ecosystem. It collects public workflows, preserves attribution, assigns trust labels, and connects interesting examples to reusable workflows.
11. botdirectory.ai (opens in a new tab)
botdirectory.ai stores reusable agent jobs as Markdown and uses GitHub as the contribution workflow. Its examples include email triage, daily briefings, SEO work and churn monitoring.
12. An open-source GTM operating system (opens in a new tab)
The grokbot-for-gtm repository packages a complete outbound workflow: account research, cold email, LinkedIn activity, reply handling and meeting logging.
13. Physical-world and home experiments (opens in a new tab)
The ecosystem is already moving outside browser-only knowledge work. The official Marketplace includes Home robots and Flora for plant care—exactly the kinds of projects worth watching because they reveal what happens when persistent agents gain narrow tools rather than being asked to “do everything.”
14. An open-source marketing OS (opens in a new tab)
Community indexes have documented a public Marketing OS built around Grok Bot with modules for audits, copy, ads, GEO, email, social, launches and positioning. This belongs in the community example bucket rather than the official-capability bucket.
15. last30days research Bot (opens in a new tab)
last30days researches what people actually say about a topic across Reddit, X, YouTube, TikTok, Hacker News, GitHub and the web—then writes a grounded brief. It is a strong Marketplace example of multi-source research with receipts.
The pattern behind the interesting builds (opens in a new tab)
The best examples do not start with “make me an AI employee.” They start with a job.
A useful Bot usually has a narrow outcome, a known set of tools, a repeatable workflow and a clear point where a person reviews or approves the result. That is also consistent with xAI’s current guidance for creating Bots.
Next: browse the full discovery list on In the wild, or jump to the 56 jobs catalogue and Getting started.