The best Grok Bot prompts

By the Chat Judge team · Updated 2026-10-19 · 5 min read · Links go to prompts in our library; creators are credited on each prompt.

In short: The Grok Bot prompts worth copying are standing instructions rather than one-off questions. They assume an assistant that keeps memory, can run routines and can read tools you connect, such as email and calendar. This guide groups the best of them and names who shared each one.

Give the assistant a job before you give it tasks

A long system prompt beats a pile of reminders. Chief of Staff system prompt, shared on X by Corey Ganim, makes the assistant the first stop for everything: it ranks your priorities, runs the calendar and task list against them, and says plainly when two things conflict and which one wins. The value is the priority list you fill in. Without it, the prompt is generic.

Speed-to-lead closer, also from Corey Ganim, is the opposite shape: a narrow role with an explicit list of what it must not do. It only turns inbound leads into booked calls, pitches only the offers on your list, and disqualifies poor fits. Narrow roles like this are easier to trust than one assistant that claims to do everything. Executive assistant for scheduling meetings, shared by Michael, sets up an assistant with its own email address that handles scheduling against rules you define, and it asks you to confirm each setup step.

Prompts that run before and after your sales calls

Pre-call brief before every sales call, shared by Michael, prepares a brief before each call on your calendar: who you are meeting, what happened last time, commitments on both sides, and open objections or risks. It tells the assistant to review your CRM, email and meeting history rather than invent context, so it is only as good as the records it can actually read.

Follow-up sweep for dormant conversations, also from Michael, looks back over recent email and calendar, finds conversations where a follow-up may have been missed, and separates people waiting on you from people you are waiting on. The instruction not to invent commitments is the important line. An assistant that fabricates a promise you never made is worse than no sweep at all.

Make the assistant learn, and audit it

Decision log that learns from your outcomes, shared by Michael, records the situation, the recommendation, what you decided and why, so later recommendations can refer to earlier ones. Personal CFO standing instructions, shared by Nikita Andriievskyi, sets a daily review of transactions with anything unusual flagged and a weekly summary. Both depend on you actually connecting the accounts they read, and on you checking the first few reports before you trust the pattern.

Workflow architect, shared by Machina, is a good corrective if you are automating too much at once. It interviews you a few questions at a time and looks for the single bottleneck that unlocks everything downstream, rather than proposing a plan that automates everything.

Then audit the setup itself. Human-dependency audit and Monthly reusable-work audit, both from Michael, look for work that stalled waiting for you and work you keep recreating. Audit every permission your bot has asks the assistant to list each permission it believes it holds, where it came from, and whether you should keep it, which is worth running after any stretch where you approved things quickly. Connect your Claude and ChatGPT subscriptions, from Corey Ganim, and Back up your bot setup, from premjith, cover wiring other models in and keeping a copy of your assistant's profile and memory somewhere you control.

Roll them out one at a time

The mistake is installing six standing prompts in a day and trusting all of them. Add one, watch it for a week, and only then add the next. Start with the chief of staff or the pre-call brief, because both produce output you can compare against your own memory of what happened.

A few more from the same people are worth knowing about. Review a shortlist against your requirements, shared by Michael, checks whether the assistant dropped options for reasons you never gave, which is a common quiet failure. Turn a good result into a reusable standard, also from Michael, extracts what made one output good and proposes an update to your instructions, so improvements accumulate instead of living in one chat. Customer promise vs delivery audit compares what your site and sales material promise with what you actually deliver. Run it before a customer does.

Briefing about you for another AI assistant, shared by Trevin Chow, and Travel-time buffers, also from Trevin and written for both Muse and Grok Bot, cover moving context between assistants and protecting your calendar. Read the briefing before you share it.

How to get good results from any of these prompts

1. Fill in every placeholder. Anything in [BRACKETS] is a fact only you know. Leave one empty and the model will guess.
2. Tell it what not to invent. Add "use only the facts above, and flag anything missing" if the prompt does not already say it.
3. Ask for one version first. Fix the direction, then ask for variations.
4. Check every name, number and promise before anything leaves your hands.
5. Edit it into your own voice. Read it aloud once; if it does not sound like you, change it.

Prompts in this guide

ResearchGrok BotChief of Staff system promptYou are [OWNER NAME]'s Chief of Staff. You are the first stop for everything. You keep [OWNER…View prompt →Corey Ganim @coreyganimReport or remove

EmailsGrok BotSpeed-to-lead closer for new inbound leadsYou are [OWNER NAME]'s Speed-to-Lead Closer. Your only job is to turn new inbound leads and warm…View prompt →Corey Ganim @coreyganimReport or remove

ResearchGrok BotExecutive assistant for scheduling meetingsYou are now my executive assistant for scheduling meetings. You have your own email address. Set…View prompt →Michael | AI + Agents @Michael_Fenech_Report or remove

ResearchGrok BotPre-call brief before every sales callBefore every sales call in my calendar, automatically prepare and send me a pre-call brief 15…View prompt →Michael | AI + Agents @Michael_Fenech_Report or remove

EmailsGrok BotFollow-up sweep for dormant conversationsReview my connected email, calendar and available meeting notes from the last 90 days. Find…View prompt →Michael | AI + Agents @Michael_Fenech_Report or remove

ResearchGrok BotDecision log that learns from your outcomesI want you to build a decision-learning system for the business decisions you help me make.…View prompt →Michael | AI + Agents @Michael_Fenech_Report or remove

ResearchGrok BotPersonal CFO standing instructionsFrom now on, you're my personal CFO. Think like Dave Ramsey: stay out of debt, spend less than I…View prompt →Nikita Andriievskyi @NAndriievskyiReport or remove

ResearchGrok BotWorkflow architect: find the one bottleneck to automateI want you to act as my workflow architect for a moment. You're someone who thinks in bottlenecks,…View prompt →MMachina @EXM7777Report or remove

ReviewsGrok BotHuman-dependency audit: where you are the bottleneckHUMAN-DEPENDENCY AUDIT Review how we’ve worked together over the last 30 days. Find every place…View prompt →Michael | AI + Agents @Michael_Fenech_Report or remove

ReviewsGrok BotMonthly reusable-work auditREUSABLE WORK AUDIT Review the work we’ve completed over the last 30 days using the conversations,…View prompt →Michael | AI + Agents @Michael_Fenech_Report or remove

ReviewsGrok BotAudit every permission your bot hasAudit every permission you currently believe you have. For each one, tell me: 1 - What action you…View prompt →Michael | AI + Agents @Michael_Fenech_Report or remove

ResearchGrok BotConnect your Claude and ChatGPT subscriptions to your botsConnect my Claude and ChatGPT subscriptions so all of my bots can use them. 1. On your computer,…View prompt →Corey Ganim @coreyganimReport or remove

ResearchGrok BotBack up your bot setup to Google DriveBack up your setup to my Google Drive. Use this folder structure, creating folders as needed: Grok…View prompt →premjith @premjithReport or remove

Keep exploring

More guides

• The best ChatGPT prompts for founders
• The best Claude prompts for founders
• The best Gemini prompts for founders
• The best Perplexity prompts for market research
• Muse prompts for your calendar and inbox
• ChatGPT vs Claude for founders’ writing
• Prompts for sales follow-ups
• Prompts for fundraising and investor updates
• Prompts for hiring your first employees

Written by the Chat Judge team from prompts in our library. We do not run model benchmarks, and nothing here is a claim about results you will get. Spot a problem? Tell us or submit a better prompt.