← Back to library

Find routines your agents run twice and merge them

Michael Fenech (@Michael_Fenech_) posted a read-only audit he ran on his own Grok Bot team after discovering two routines doing the same job. It compares what each routine actually does rather than its name. We explain when to run it and show his prompt word for word with credit.

  • OutcomeHave one agent collect every scheduled routine across your agent team, compare the saved instructions of any that look alike, flag pairs that read identical sources or run too close together, and hand you a short keep, merge or cut table with the expected saving, without changing anything until you say yes.
  • SetupMinutes
  • LevelStarter
  • CreditMichael Fenech (@Michael_Fenech_)
View original
Works with
Grok BotClaude CoworkChatGPT agentCustom agent

Published Nov 2, 2026Updated Nov 2, 2026AI dry run on sample input: passed Nov 2, 2026Source verified Nov 2, 2026

When should you use it?

  • Several agents were set up at different times and some of their jobs sound alike
  • Two morning summaries or two inbox checks arrive within minutes of each other
  • Your included usage runs out sooner than the work seems to justify

Not for: A single agent with one or two routines you can check by eye; Platforms where agents cannot message each other or read saved routines; Deciding which agent should own a job when the overlap is deliberate backup

What do you need, and what do you get?

You need

  • A lead agent that can list its own routines and message the others
  • Your time zone, so schedules line up in one table
  • Optionally, your usage report to sanity-check the claimed savings

You get

  • One table of every routine with its owner, schedule and purpose
  • Pairs flagged for pulling the same data or checking the same thing close together
  • Routines flagged for running more often than their job needs
  • A keep, merge or cut recommendation for each, with the saving
  • A note of any agent that did not reply and is therefore unverified

How does it work, step by step?

  1. Ask for the inventoryPaste the prompt into your lead agent; it lists its own routines and asks the others for theirs.
  2. Chase silent agentsAsk any agent marked unverified directly, so the table is complete.
  3. Read the overlapsOpen the paired instructions yourself for anything flagged as a merge.
  4. Merge one pairApprove a single merge, keep the old routine paused rather than deleted, and compare a week of output.
  5. Lower frequenciesCut the run rate of routines flagged as too frequent, starting with the costliest.

The prompt

The original prompt by Michael Fenech (@Michael_Fenech_), word for word, with a link to where it was published. Below it is our adapted version with fill-in blanks, if you want it. Works in Grok Bot, Claude Cowork, ChatGPT agent and others. Values in lime are placeholders.

Original prompt by Michael Fenech (@Michael_Fenech_)

Word for wordby Michael Fenech (@Michael_Fenech_)
Audit my Grok Bot team for duplicate routines. Read-only: change nothing.

1 - List every routine you have: name, schedule (my time zone), and one line on what it checks or produces.

2 - Message each of my other bots and ask them to reply with the same list. Tell them to make no changes. Mark any bot that doesn't answer as unverified.

3 - For any routines that look similar, open their saved instructions and compare what they actually do, not just their names.

4 - Flag routines that pull the same data, check the same thing, or run within an hour of each other on the same source.

5 - Also flag any routine that runs more often than its job needs.

Give me one short table: the routines, which bot owns each, what overlaps, and your recommendation (keep, merge or cut) with the saving. Don't change anything until I say yes.

Exactly as published by Michael Fenech (@Michael_Fenech_) in X post: a read-only audit for duplicate Grok Bot routines.

Our adapted version (optional)

Fill in the blanks (optional; the copied prompt uses your values)

Adapted prompt2 variables
Check my agents on {platform} for scheduled jobs that repeat each other. Do not change, pause or create anything.

Build one table of all routines across every agent: owner, schedule in {time_zone}, and what each one produces. Ask the other agents for their lists, and mark any that do not answer as unknown.
Where two routines look alike, compare their saved instructions and say whether they share a source, overlap in purpose or are scheduled close together.
Also point out any routine scheduled more frequently than its source changes.
For each, recommend keep, merge or cut, estimate the usage saved, and wait for my approval.

What can go wrong?

  • Two routines with the same source can still serve different readers. Check who uses each output before merging.
  • Pause before you delete, so you can restore a routine if the merged version misses something.
  • Savings estimates are guesses until you compare a week of usage before and after.
  • An agent that does not answer is not proof it has no routines; treat it as unknown.

Our take

Duplicate routines are an easy mistake to make once you have more than a couple of agents, because each one was set up for a good reason at a different time, and the overlap only shows when you compare what they do. Fenech's prompt is short, and that is its strength: it is read-only from the first line, it compares saved instructions instead of names, and it marks any agent that stays silent as unverified rather than assuming it is clean. The frequency check is a nice extra, since a routine running hourly when the data changes daily wastes as much as a duplicate. It pairs well with a single-routine simplification review: run this one first to remove whole repeats, then trim the steps of what remains. Our caution is to pause rather than delete when you merge, and to confirm who reads each output, since two similar reports may serve different people. In our dry run on a fictional three-agent sales team, it found two inbox summaries 20 minutes apart and a competitor scan running hourly against a weekly blog, and recommended one merge and one cut.

Where the time goes back: You stop paying twice for the same check, get one clear summary instead of two near-copies, and free up usage for work that is actually new.

FAQ

Will it delete anything?
It should not. Fenech's prompt is explicitly read-only and ends by asking for a yes before any change.
How is this different from simplifying one routine?
A simplification audit shortens the steps inside one job. This one looks across agents for whole jobs that repeat each other or run too often.
Can I use it with Claude Cowork or ChatGPT?
Yes, as long as one assistant can see or ask about the scheduled tasks of the others. If it cannot, list them yourself and paste them in.

Source and credit

The original prompt on this page is by Michael Fenech (@Michael_Fenech_), shown word for word with a link to where it was published. The summary, notes and adapted version are ours.

Original
X post: a read-only audit for duplicate Grok Bot routines
Creator
Michael Fenech (@Michael_Fenech_)
Published on
X
Original date
Nov 1, 2026
License
Not openly licensed; the original prompt is shown word for word with credit and a link to the original
Originally for
Grok Bot
AI dry run
Nov 2, 2026: passed. Dry-run tested by our AI test agent on sample input. Compared instructions not names, found the duplicate and the over-frequent routine, gave keep/merge/cut with savings, read-only.
AI dry run on sample data; not run with live connected accounts. The playbooks.bot test agent filled the prompt with synthetic inputs, produced the output and checked it against the page's stated outputs and guardrails.
Verified
Nov 2, 2026: source link resolves to the original, the credited creator published it, our excerpt or prompt matches it, and the license allows our use.

Are you the creator? Claim or correct this page.

More in Operations & Admin

All 203 Operations & Admin playbooks →

Related playbooks

Operations & AdminPrompt✓ AI dry run

Audit how you manage your agent team, not only the agents

Have your chief-of-staff agent review the last 30 days of how you delegated, clarified, interrupted, corrected, assigned, approved and followed up across your agent team, score eight management areas with evidence, list your three biggest weaknesses, and draft a seven-day improvement plan, without changing any bot, routine or instructions until you approve.

Credit: Michael Fenech (@Michael_Fenech_)

Under an hour
Operations & AdminPrompt✓ AI dry run

Brief an agent with nine fields, not a how-to

Before you hand a job to an agent, fill in nine short fields that name the result, the sources it may use, the deadline, what it may do alone, what needs your say-so, what it must never do, how to prove the work is finished, what to do when it gets stuck, and where to put the deliverable, then add a line that tells it to pick the simplest reliable method.

Credit: Michael Fenech (@Michael_Fenech_)

Minutes
Operations & AdminPrompt✓ AI dry run

Brief your agent on the outcome, not every step

Hand a new job to your lead agent by describing the finished result instead of the steps, and have it confirm what success looks like, ask only for context it cannot find itself, pick the simplest dependable approach, state what it will do alone and what needs your approval, and check the result against your goal before calling it done.

Credit: Michael Fenech (@Michael_Fenech_)

Minutes
Operations & AdminPrompt✓ AI dry run

Give every agent routine a clear point where it stops

Have your lead agent review every recurring job, monitor and follow-up your agents run, check each against four stop rules (what counts as done, what should hand control back to you, how many retries or days are too many, and how a stop gets recorded), then propose fixes ranked by risk, changing nothing until you approve.

Credit: Michael Fenech (@Michael_Fenech_)

Under an hour
Operations & AdminPrompt✓ AI dry run

Have your agent trace the ripple effects of each decision

Add a standing rule to your project agent so that whenever you confirm a decision, it searches the plan, email, calendar and documents for the tasks, dates and promises that depended on the old choice, updates what it is allowed to change, drafts the rest for your approval and reports what moved and why.

Credit: Michael Fenech (@Michael_Fenech_)

Minutes