Hire an AI reversible decision experiment planner

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• A sort of the choice as a one-way or two-way door, with the process that fits it
• A long-shot plan with the small test, the most you can lose, one owner and a rule for dropping it
• A review of each program, marked keep, cut or double down, with the reason

How it does the job well

• Start from the user problem and the evidence for it, then the solution.
• Write decisions down with the reason and the options you rejected.
• Keep scope small enough to ship and learn. Name what is out of scope.
• Define how success will be measured before building.

Mistakes it avoids

• Treating opinions or one loud request as proven demand.
• Specs that skip edge cases, error states or dependencies.
• Roadmaps with dates but no owners or confidence levels.

What a good result looks like

An example, not your task: You are stuck on a choice and can't tell whether it needs a week of debate or an hour.

• Step 1: Decision sort. The choice marked as a one-way or two-way door, with the owner, review date and what is still needed.
• Step 2: Long-shot plan. The bet, honest odds and payoff, the most you can lose, one owner, a short build and what counts as learning.
• Step 3: Keep-or-cut review. Each program with its cost, return and customer signals, marked keep, cut or double down, with the reason.
• Hand-back summary. Each step's file, any step not run and why, and every MISSING line in one list.

How the work is checked

• The sort names what makes the choice hard or easy to undo.
• A two-way choice has one owner and a review date. A one-way choice lists what is still unknown.
• The long shot has a loss cap, a short build time and a stated rule for what counts as learning. The odds are honest, not a hopeful guess.
• Every program in the review has a decision and a reason, and cash comes before how it looks on paper.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

A second opinion

Before it hands back, your AI reviews the draft the way Jeff Bezos and Warren Buffett would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Sort the decision. States the choice and asks what would be hard to undo. Sorts it as a one-way or two-way door and picks the process that fits.
• Plan the long shot. Takes the sorted choice and plans a small test. Sets the odds, the payoff, the most you can lose and what counts as learning if it fails.
• Review each program. After your test, lists each program with its cost and return. Checks customer signals before the numbers, and marks each keep, cut or double down.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id reversible-decisions only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/reversible-decisions/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• The choice you are stuck on, and who it touches
• A long-shot bet you want to try, and what a big win would look like
• A sheet of what each of your programs cost and what it returned, for the last step
• Customer signals for each program, such as repeat purchases and referrals (optional)

Helpful context (optional)

• Cost and return for each program
• Repeat purchase and referral numbers
• A note of who would object to the choice

Tools and access

• Any good chat AI: Holds the choice, the plan and your results across the pause.
• A spreadsheet: Tracks the cost and results of the small test and each program.
• Any good chat AI: Sorting a choice and writing the process is text work. Claude, ChatGPT and Gemini all handle it with this skill.
• A shared document: The written call and dissent stay where the whole team can see them.
• Any good chat AI: Setting limits and listing tests is text work any strong chat AI handles with this skill.
• A spreadsheet: A simple sheet shows cost against the loss cap as the tests run.
• Any good chat AI: Comparing lines and writing verdicts is text and table work that any strong chat AI does with this skill.
• A spreadsheet: The skill reads tables best, and you can check its sums against yours.

Functions and deliverables

• Reversible decision experiment planner: Share the choice, risks and resource limits. Get an undo-cost check, small experiment, loss cap and later keep-or-stop review.
  • A sort of the choice as a one-way or two-way door, with the process that fits it
  • A long-shot plan with the small test, the most you can lose, one owner and a rule for dropping it
  • A review of each program, marked keep, cut or double down, with the reason
• Reversible decision sorter: Sort a stuck decision as a one-way or two-way door. Get the review it needs, plus an owner, a small trial and a review date.
  • The choice sorted as a one-way or two-way door, with the reason
  • The cost of reversing it, in time, money and trust
  • The review that fits: wide and slow, or small and fast
  • For a two-way door, an owner, a small trial and a review date
  • For a one-way door, what to learn first, and whether a smaller step would make it two-way
  • A written call with any dissent on record
• Long-shot bet planner: Test a risky idea without risking the business. Get the odds and payoff, a loss cap you can survive and a rule for dropping it.
  • The odds and payoff written down, with a check that the upside is not capped
  • A loss cap you can survive
  • Cheap tests, cheapest first, each with its cost and what counts as a pass
  • A decision date, a measure and rules for dropping the bet or going bigger
• Program keep-or-cut reviewer: Paste results for each project you run. This program keep-or-cut review marks every one keep, cut or double down, judged on cash and customers.
  • A verdict for each line, either keep, cut or double down
  • Earnings and new cash needed side by side
  • The four tests for each line, and a note on what customers are doing
  • A lesson for every cut line, and a reason short enough to share
• Product context file writer: Share product facts, audience and positioning. Get a reviewed context file that other marketing skills can reference.
  • A product marketing context file
  • A summary of what the file already holds
  • A change log of updates
  • Gaps you still need to fill
  Source: Corey Haines.
• Jeff Bezos: Review a product or stuck decision through Jeff Bezos’s recorded approach. Start with the customer, test what can be undone and write the reasoning.
  • Your problem worked on the way Jeff Bezos would: their questions, their method, their push-back
  • The answer or the work, in Jeff Bezos's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step
• Warren Buffett: Review a business using Warren Buffett’s recorded principles. Check how it earns, what protects it, how managers use cash and which facts remain uncertain.
  • Your problem worked on the way Warren Buffett would: their questions, their method, their push-back
  • The answer or the work, in Warren Buffett's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step

Tasks you can use it for

• You are stuck on a choice and can't tell whether it needs a week of debate or an hour.
• You have an idea with a low chance and a big payoff, and you want to try it without risking the business.
• You run several programs, such as channels or products, and need to decide which to keep, cut or push.

Limits to check before starting

• You face a daily operational call: use a normal review meeting or ask the person who owns that task.
• You need to rank many big choices by importance: use the Decision priority planner.
• You want a group to weigh in on a choice: use the Group decision reviewer.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Download the handoff pack. It's a zip with HANDOFF.md and every skill file this workflow uses.
• Give the folder to Claude (in Claude Code, or as files in a Claude Project) and say "Follow HANDOFF.md".
• Paste the choice you are stuck on. It sorts it and plans the long-shot test.
• It stops after the plan. Run the small test and write down its cost and results.
• Bring the results and your program numbers back and say "carry on from step 3".

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/reversible-decisions for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.