GrokBot

Message a team of AI agents on desktop + iOS — shared plugins/skills, cloud computers, agent-to-agent work, and Zapier-like triggers — aimed at Claude Cowork / ChatGPT Work territory

GrokBot table of contents  ·  ← AI Agents  ·  AI  ·  Productivity  ·  ChatGPT Work  ·  God agent vs team  ·  Grok Build  ·  Buzz  ·  exe.dev  ·  GrokBot CEO pattern  ·  GrokBot social  ·  GrokBot course  ·  GrokBot hedge fund  ·  GrokBot Elon  ·  GrokBot X  ·  GrokBot vs Hermes  ·  GrokBot tips  ·  GrokBot field manual  ·  GrokBot SEO agent  ·  GrokBot limits  ·  GrokBot team

Sources

Riley Brown — product walkthrough

Riley Brown (@rileybrown) — Agent Native / agents in business; podcast @agentnative_.

x.com/rileybrown/status/2087235887012749383 (11 Aug 2026) — create agents + connections → plugins → skills/descriptions → cloud computer → agents talking → triggers → what’s missing. Riley follow-up: group chats work (correction).

Official intro + 100 use cases

Official beta post @bot / Introducing Grok Bot — “AI teammates that do real work… sign in to your tools… come back with finished work.”

100 use cases compiled by eric zakariasson (@ericzakariasson, Cursor) — x.com/i/status/2087258914060664902 (11 Aug 2026). High signal (~2k bookmarks at capture). Full list + themes below; Mike-priority picks mapped to this portfolio.

One-sentence TL;DR

GrokBot is a desktop + iOS multi-agent workspace: you create named specialists (EA, content, developer…), connect tools via chat or a plugins tab, give each agent a cloud computer, let agents message each other, and run time-based routines + event triggers (Slack, Git, Linear, etc.) — synced so phone and desktop stay one system.

What you get (feature map)

Area What Riley shows
Surfaces Desktop app + iOS app, fully in sync (message on phone, see it on desktop)
Agents Create bots with name, title, description = standing system prompt / personality
Connections Authorize plugins in conversation (e.g. Linear, Slack) or via Plugins tab
Plugins vs skills Plugins @-mentionable (Gmail, Vercel, ReMotion…); skills via slash (e.g. Scrape Creators) — Codex/Claude Code–like split
Shared library Plugins and skills are global across agents; descriptions differentiate behavior
Cloud computer Each agent has a full remote desktop (Chrome, localhost, internet); control from phone too
A2A Agents ping each other; human sees meta thread + full agent↔agent transcript
Routines Create by talking (“every day at 9am…”) → shows in side panel
Triggers Event-driven: Slack channel, Git event, Teams, Linear, Sentry, PagerDuty (more expected)
Nav Cmd+K search agents/messages; links shared in chats; files UI not solid yet

Walkthrough highlights

1. Create a bot + connect tools in chat

Plus → new bot animation → “what do you mainly want me around for?” Example: email + calendar manager, also Slack + Linear. Agent pulls plugins; you authorize in-product. Then: “At 9am, morning digest: Linear, Slack, email + calendar…” → becomes a routine in the side panel (chat / computer / routines / settings).

2. Description = the real differentiator

Name + title + description is the custom agent prompt. Content agent vs developer agent only behave differently because of those strings — not because plugins are siloed. Implication: invest in descriptions the way you invest in soul.md / AGENTS.md / company brain briefs.

3. Cloud computer for the agent (not for you to browse ESPN)

Full remote desktop: agent builds a landing page, runs it on localhost:3001, screenshots progress. iOS can open the same computer. Point is execution environment for the agent — same spirit as exe.dev VMs and ChatGPT Work cloud computer, productized per agent.

4. Agents talk to each other

Developer agent voice-prompted from phone: talk to content agent, use scraped Instagram transcripts → app idea → build iOS/Android prototype → deploy (Revel-style simulator). Human gets pings from both; can open the full agent↔agent thread. Animations show which agents are “working.”

5. Triggers = Zapier with an agent brain

Time routines and event triggers. Demo: Slack channel short-form-videos → agent summarizes. At capture, trigger apps were limited (Slack, Git, Teams, Linear, Sentry, PagerDuty); Riley expects email + more soon.

What’s missing (Riley’s list) / likely soon

  • GrokBot as a native Slack bot you @ in channels — Cursor bot already works in Slack; Riley expects GrokBot bots to follow (platform lineage).
  • Files tab — create files in chat, not reliably showing in Files UI yet.
  • More triggers — email and broader app graph.
  • Groups — UI clues + agent claims; Riley later confirmed group chats work.

Riley comparison: feels like an easier cloud Buzz (Jack Dorsey’s agent-friendly Slack-like) where every agent brings a cloud computer.

100 use cases (Eric Zakariasson)

Beta reality check: bots sign into tools, drive GUIs when APIs are missing, run fleets, and ask before spending. Grouped for scanning — numbering matches Eric’s thread.

Themes at a glance

GTM & sales (1–27)

Personalized outbound, SQL→copy→reverse ETL, multi-bot GTM crews, CRM rebuilds, event ICP screening, SFDC from phone, Gong win/loss, LinkedIn mapping, QBRs, churn signals.

Ads, content, enablement (28–40)

Search-ads pacing, creative tests, competitor price watch, Figma ads, voice-matched posts, follow-ups, demo clip libraries, security questionnaires, “what did we promise?” packs.

Eng / data / fleet ops (41–50)

Datadog flamegraphs, warehouse→charts, migration ramps, cloud IDE/SSH, Vercel prototypes, model evals, conflicting PRs, tweet→CSV, browser bots when no MCP, kill wasteful bots.

EA, Slack, onboarding (51–67)

Channel PA as you, auto-propose automations, morning digests, focus time, email triage, new-hire week-one plans, phone/subway deck edits, gym texting, laptop janitor.

Support & money (68–72, 81–86)

Expense photos, invoices, support queue hourly pass, tickets→changelog/help center, subscriptions cancel, finance narrative, insurance claims, lease diffs, charge disputes.

Life / consumer ops (73–100)

Campsites, tickets, gym pool snipes, groceries from fridge photos, Amazon cart, travel points, Anki from Kindle, school forms, passport slots, wedding digest — same “finished work” pattern as B2B.

Highest leverage for Mike / Shadstone / Para Living

# Use case (Eric) Why it fits us
3–4 Multi-bot GTM crew; chief-of-staff hub → specialist fleet Same shape as god agent vs team + 5 roles
6–7, 10 Community ops fleet; event ICP screen; Luma+DB+Slack weekly summary GFAVIP / CBS / events; keep rooms intentional
8 Bot for redemptions with human approval before spend Wallet / perks / GFAVIP money paths — never silent spend
16–17, 53 Company chat search; daily brief per meeting person; multi-inbox morning digest EA bot week one on GrokBot
30–32 Competitor pricing watch; ads tests backlog; voice-matched LinkedIn/X drafts Product brain + content agent; human approve posts
39–40 Security questionnaire draft; “what did we promise this customer?” Partner / enterprise deals; contracts + Slack + CRM on one page
44–47 Cloud workspace + Bot; VM → screenshot + Vercel; evals; flag conflicting PRs Pairs with exe.dev + Grok Build for real shipping
50–52 Vet fleet for spend; Slack PA; propose automations from history Stop bot sprawl; compound workflows without re-briefing
70–71 Support queue drafts; tickets → changelog + help articles Shadstone support / product ops loop
77 Snipe always-booked gym pool spots Funny twin of gymhack — use allowlists + human confirm, not silent cancel

Full list (1–100)

Expand full numbered list from Eric’s thread
  1. Book meetings via personalized outbound in your Gmail voice, then send the notes
  2. Sales play that was 90 min–2 hr (SQL audience → copy → reverse ETL) → ~95% automated
  3. Multi-bot GTM crew (research, outbound, forecast, inbox, slides) + weekly playbooks
  4. Chief-of-staff hub that routes work to a large specialist fleet
  5. Rebuild most of a custom CRM with a small bot crew in ~1.5 days
  6. Community ops fleet that saves 20+ hours a week
  7. Screen 1000+ event applicants against an ICP; batch-approve fits
  8. Replace pricey direct-mail SaaS; bot handles redemptions and asks before it spends
  9. Review ~200 recruiting applications into strong / mid / reject in one pass
  10. Weekly Luma + database + Slack summary; free 2–3 hours/week
  11. Update a sales deck live from Granola notes mid-call (stop recording early)
  12. Localize a sales deck before a customer call
  13. Answer a Salesforce EOQ question from your phone in ~10s (no laptop / SFDC app)
  14. Update a Salesforce org chart from just a name or email
  15. Build a Salesforce report + dashboard so you stop re-asking opp questions
  16. Search company chat for who asked about a feature in ~15s
  17. Daily brief with usage data for every person you’re talking to that day
  18. Coach yourself on Gong calls with timestamped comments + homework
  19. Week of Gong calls → win/loss memo with phrases that closed
  20. Call questions → living Notion FAQ updated every morning
  21. LinkedIn watch for eng/product/AI leader events → weekday digest of ready-to-send notes in your voice
  22. Personalized LinkedIn connection requests from Sales Navigator; book eng-leader meetings
  23. Map an account from LinkedIn + CRM; show who influences the deal
  24. Audit/rewrite LinkedIn profile from live page + Notion + Slack; human approval before publish
  25. Forecast account health / catch churn when usage looks fine but is dropping
  26. Tier accounts (fit × warmth) Salesforce → Notion; batch enrich contacts
  27. QBR pack: usage + tickets + open opps + short slide narrative
  28. Daily Apple Search Ads pacing digest; reply in-thread to edit bids
  29. Scan ad performance; explain winners; suggest tests into a Notion backlog
  30. Watch a competitor’s pricing page; Slack when something changes
  31. First-pass performance ads in Figma matching existing format
  32. Learn writing voice from past posts; draft LinkedIn/X into Notion for review
  33. 30 min after a call: follow-up reminders (optional character voice)
  34. Work around missing Outreach bulk-enroll by driving the GUI
  35. Record yourself once doing a reply/SFDC workflow → bot turns it into automation
  36. Sync local marketing calendar from global webinars Notion in ~2 minutes
  37. Customer enablement pack end-to-end: find request email, Zoom downloads, summarize, Drive, draft reply
  38. Recorded demo → tagged clip library for the next demo
  39. Draft security questionnaire from public docs; flag human-only questions
  40. “What did we promise this customer?” from contracts + Slack + CRM on one page
  41. Log into Datadog; spin bots to read flamegraphs when profiling API isn’t exposed
  42. Pull warehouse data; build an animated chart in tldraw
  43. Monitor a DB migration ramp; check logs/metrics; walk the backfill
  44. Spin up a cloud workspace with IDE + SSH; bot working in minutes
  45. Prototype on a VM; get a screenshot + Vercel link back
  46. Run an eval suite across models; compare quality vs cost; pick what to ship
  47. Flag conflicting pull requests across a busy repo
  48. Scrape a tweet topic into a CSV with the columns you want, one shot
  49. Browser bot stays logged into iCloud so other bots reach contacts/reminders when there’s no MCP
  50. Vet the whole bot fleet for spend; kill the wasteful ones
  51. Personal assistant on a Slack channel that answers as you
  52. Inspect Slack/Notion/GitHub/Linear history; propose two automations without you describing the job
  53. Early-morning digest across Slack, Gmail, calendar, Granola; manage todos from a DM
  54. Find 90 minutes of focus time this week; decline junk meetings
  55. Triage email twice a day
  56. Filter noisy Slack: only onboarding-relevant actionable pings with context + links
  57. Organize, prune, and join the right Slack channels
  58. Rebuild notes system (Obsidian + Granola + Gong + calendar) in ~10 minutes week one of a new job
  59. First-week plan for a new hire from last few people’s calendar + Slack
  60. Rebuild a messy Drive folder into a wiki with owners and stale banners
  61. Update decks from your phone on the subway
  62. Meeting prep + decks pushed before you get to the office
  63. Daily Slack digest → ~3-minute morning podcast on mobile
  64. Text your bot from the gym and keep working
  65. Trigger laptop actions from phone even when desktop app is closed
  66. Clean GB of installers/junk; daily evening laptop janitor
  67. Monday-morning batch of personal Slack messages in a fraction of the time
  68. Track invoice actuals for big expenses (bookkeeping-style skill)
  69. File expenses from a photo dump of receipts, coded to category
  70. Hourly pass on support queue; draft replies; only ping when human needed
  71. Week of support tickets → changelog notes + help-center articles
  72. Monitor incident-related volume in the background
  73. Reconstruct dates + approx GPS for analog film TIFFs from folders, receipts, Photos library
  74. Guide a house paint negotiation (and contractor TikToks for color); save thousands
  75. Monitor recreation / ReserveCalifornia for campsites; book after confirm
  76. Watch an AMC ticket drop; buy when seats open
  77. Snipe always-booked gym pool spots
  78. Price-compare Instacart vs Amazon; place the grocery order
  79. Recipe photo or IG reel → grocery cart with smart substitutions
  80. Week of dinners from a fridge photo + store delivery window
  81. Gmail receipts → subscription list; cancel forgotten ones; unsubscribe newsletters
  82. Personal finance web app: clean rules; narrative money summary
  83. Assemble an insurance claim: photos, receipts, timeline, email to send
  84. Compare two lease or insurance PDFs into a one-pager of what changes
  85. Tax PDFs → labeled folder + questions for accountant
  86. Dispute a wrong charge: receipt, letter draft, track case number
  87. Find exact products matching criteria in ~20s (bike racks, tickets, shoes, tires…)
  88. Amazon cart with gift wrap + ship-to; place order without touching checkout
  89. Explore Chase points redemptions across dates and cities
  90. Fantasy football dynasty draft cheat sheet from current roster
  91. Find Airbnbs; build a slide deck; email partner
  92. Kindle highlights → Anki cards every morning
  93. Kids activities that fit ages and nap schedules in ~30s
  94. Fill school/camp forms from last year’s PDFs; only ping for signatures
  95. Scrape/parse/transpose piano practice PDFs into the format you use
  96. Government license / IDP process checklist while you’re at the office
  97. Check into flights when window opens; only ping for 2FA or captcha
  98. Watch passport/visa/DMV appointment slots; grab after confirm
  99. Redesign personal portfolio in Webflow with real back-and-forth on custom code
  100. Plan a multi-day wedding with vendors + Sunday digest

Pattern across almost all 100: finished work artifacts (deck, CSV, draft, booking, report), often with human confirm on money or publish, not “another chat answer.”

GrokBot vs Grok Build (how we already work)

You like Grok and Grok Build (terminal TUI, multi-thread SOP, HEOS-style repos — /grokbuild, /grokthreads, /grok-mobile). They are not the same product:

Grok Build (what we use heavily) GrokBot (this launch)
Shape CLI / TUI coding agent in a repo Multi-agent messaging app (desktop + iOS)
Best for Code, site pages, git, structured engineering sessions Knowledge work fleet: EA, content, ops, cross-tool routines
Compute Your machine / SSH box / optional remote Per-agent cloud computer + plugins
Multi-agent You orchestrate threads / worktrees Agents message each other natively
Mobile SSH thin client (notes) Native iOS, full sync

Recommended split for Mike

Keep Grok Build as the deep coding / site-shipping cockpit (Astro pages, HEOS, RankingSolution code). Use GrokBot as the always-on team of specialists for email/calendar, content ops, Slack triggers, and light product prototypes — hand hard engineering back to Grok Build or Codex when the repo is the product.

How you should use GrokBot (Mike-specific)

Matches how you already think: one front door + specialists, 5 roles, company brain, multi-company portfolio (Para Living, GFAVIP, Headless Domains, PowerLobster, this site).

Next level: once you have specialists, put a CEO / chief-of-staff bot on top so you stop being the middleman between every chat (Teslaconomics pattern). The official product walkthrough of that desk is the GrokBot course (Amrita workshop).

Starter fleet (create these first)

1. Chief of staff / EA

Email + calendar + Slack digest. Morning routine: calendar first, then sponsorship / partner emails, then tickets. Description: ruthless triage, never invent commitments. Links: send and delete, chief of staff.

2. Content agent

Blog + X + video notes. Skills: scrape creators / social transcripts (Scrape Creators). Description: Mike’s voice (see brand-voice skill), field-note structure like this site, always cite sources. Can draft *.astro outlines; ship via Grok Build.

3. Developer agent

Prototypes on cloud computer (landing pages, small tools). Description: prefer Astro, don’t invent APIs, hand production PRs to Grok Build / Codex. Vercel/Cloudflare plugins when demos need a URL.

4. Product brain agent

Para Living (paraliving.com) + theflysales.com research. Description: every insight tagged to a product/SKU; competitors, market, listings, reviews — /productbrain method.

5. SEO / Ranking agent

GSC / SERP / RankingSolution loop language from /seoloop. Trigger on Linear issues or Git if SEO tasks live there; weekly routine for opportunity briefs.

6. Ops / community agent

GFAVIP / PowerLobster / Slack channels: summarize, never auto-post sensitive stuff. Triggers on specific Slack rooms (short-form, partner, support). Aligns with PowerLobster privacy toggles.

High-value automations to set in week one

Mapped to Eric’s list numbers where useful:

  1. 09:00 digest (EA) — #53 style: calendar → urgent Slack → Linear/inbox themes.
  2. Slack short-form / content channel trigger — #32: draft X/LinkedIn in voice; human approves posts.
  3. Competitor / product pulse — #30 + product brain: pricing page watch + weekly SKU brief for theflysales / Para Living.
  4. Git / PR hygiene — #47: flag conflicting PRs; merge still human + Grok Build.
  5. A2A handoff — #3–4 + #45: content → developer prototype (VM/screenshot/Vercel); production Astro pages stay Grok Build.
  6. Fleet spend review — #50: weekly pass to kill wasteful bots before they multiply.

Plugins / skills to prioritize

  • Slack, Linear, Gmail/calendar — EA + ops skeleton
  • Vercel / Cloudflare — demo deploys
  • Scrape Creators–class skill — content corpus (Riley’s favorite pattern)
  • Remotion / video stack only if you’re actively shipping video automation
  • Avoid dumping bank / production admin into every agent — shared plugins mean description discipline + least privilege (see gymhack)

Guardrails (same lessons as our other agent pages)

  • Shared plugins = shared blast radius. One Gmail plugin is available to all bots. Use descriptions + human approval for send/delete/spend.
  • Cloud computer ≠ production laptop. Prototype there; ship production code through git + Grok Build / CI.
  • Triggers can spam. Start with one Slack channel and a summary-only instruction.
  • Don’t abandon Grok Build. GrokBot is the fleet UI; Grok Build is still the best place for deep, multi-thread engineering on your real repos.

30-day experiment plan

Week Focus
1 EA + Content only; morning digest; one Slack trigger; polish descriptions
2 Product brain agent for one Para Living line; weekly routine; store briefs where theflysales expects
3 Developer agent for demos only; one A2A content→dev prototype; no production deploys unattended
4 SEO/ops agent; review which routines actually saved time; kill noisy triggers

Related on this site

Sources: Riley Brown walkthrough · Official Grok Bot intro · Eric Zakariasson — 100 use cases

Field notes · August 2026 · GrokBot for the fleet · Grok Build for the repo · 100 use cases · @rileybrown

Comments

Approved comments appear below. Log in once with GFAVIP — it applies across the whole site. GFAVIP login

View comments archive