Hire an AI decision process reviewer

Everything it can hand back

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

• The decision sorted as easy or hard to undo, with the cost of undoing it
• How much evidence is enough, what to get first and a date to decide
• Your checklist answered item by item, with a clear go, change, wait or not-ready call
• A review of the process later, and changes to your checklist for you to approve

How it does the job well

• Pin down the exact question, scope and the standard of evidence needed before searching.
• Prefer primary sources, and say how strong each source is.
• Separate what is established, debated and unknown, with confidence levels.
• Cite specifically: author, work, date or page where possible.

Mistakes it avoids

• Calling a decision good because it turned out well, or bad because it did not.
• Adding a checklist item for every miss. Cut as often as you add. Twelve items is the limit.
• Running the heavy checks on two-way doors.
• Looping more than twice on the checklist.
• Presenting one source or a popular claim as settled.
• Inventing citations, quotes or page numbers.

What a good result looks like

An example, not your task: You face a decision that is hard to undo and want it checked the same way each time.

• Decision and door. One line on the choice, whether it is easy or hard to undo, and what undoing would cost.
• Evidence bar. How much evidence is enough, how much you have, what to get first and the date to decide.
• Pressures and the case against. The strongest pressures and the best argument against the choice, for hard-to-undo decisions only.
• Ruin check. The outcome you could not recover from and a capped version, when money, time or your name is at stake.
• Checklist result. Each checklist item marked pass, fail or unknown, with evidence and the number of rounds it took.
• The call. Go, change, wait or not ready, with the reason, what would prove it wrong and a review date.

How the work is checked

• The door was sorted before any other check ran.
• A one-way door got the pressure check and the case against.
• No "go" was given while a stop item failed or the ruin check had an open unrecoverable outcome.
• The checklist loop ran at most twice.
• 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 Charlie Munger and Jeff Bezos would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Sort the door. Asks how hard the choice is to undo and what undoing would cost. That sets how much checking it gets.
• Set the evidence bar. Says how much evidence is enough, what to get first and a date to decide.
• Choose the checks. An easy-to-undo choice gets a light check. A hard-to-undo one gets the pressure check and the best case against it. If money, time or your name is at stake, it also tests what could ruin you.
• Run your checklist. Answers each item with evidence. If a stop item fails, or three or more items are unknown, it asks for facts and runs again, at most twice.
• Record the call. Saves the call, the reasons, what would prove it wrong and a review date. You make the decision.
• Review and learn. On the review date, it judges the process, writes a rule if a weak check caused a bad result, and proposes changes to your list.

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 decision-checklist-keeper 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/decision-checklist-keeper/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 decision, who it touches and when it is due
• What you know and what you wish you knew
• What is at stake, for choices that are hard to undo
• Your checklist and past decisions, if you have them (optional)

Helpful context (optional)

• Notes from past decisions and how they turned out
• The numbers behind the choice, such as costs, dates and quotes

Tools and access

• Claude Code: It can keep the decisions folder and read the checklist and log each time.
• Notion or Google Docs: Keep your decision log here and paste a row in when a review date comes.
• 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: Weighting facts and doing the share sum is text work any strong chat AI does well with this skill.
• Any good chat AI: It is a reasoning task on text that any strong chat AI can do with this skill.
• A trusted second reader: A person outside the situation can spot pressures you cannot see from inside.
• Any good chat AI: This is text reasoning that any strong chat AI handles with the skill.
• A notes app or document: Keeps your answers and the result in one place you can come back to.
• Any good chat AI: Listing dependencies and testing plans in text is what any strong chat AI does with this skill.
• A spreadsheet: Handy for putting fixed costs, cash and exposure side by side.
• A notes app or document: Keeps the log, the checklist and your records in one place you can come back to.
• Any good chat AI: Structuring decisions and writing plans is text work any strong chat AI does well with this skill.
• A notes app or document: Gives you one place to keep rules and the dates to check them.

Functions and deliverables

• Decision process reviewer: Check a major choice before you act. Record the reasons, then review the process when the result is known.
  • The decision sorted as easy or hard to undo, with the cost of undoing it
  • How much evidence is enough, what to get first and a date to decide
  • Your checklist answered item by item, with a clear go, change, wait or not-ready call
  • A review of the process later, and changes to your checklist for you to approve
• 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
• Decision evidence planner: Find out how much evidence is enough before you decide. Paste your decision and what you know. Get a decide now or wait call and a date.
  • A weighted list of facts, marked known, partly known or unknown
  • Your information share as a percent
  • The bar for this decision
  • A decide now or wait verdict with the reasoning
  • The one fact to get first and a call date
• Decision bias checker: Check a decision for pressure before you commit. Describe it and get the pressures bending your judgment, plus a calmer version of the choice.
  • Your rational case, written before the pressure check
  • The pressures present, each tied to a fact in your situation
  • A flag when three or more stack, with a cooling-off step
  • A calmer version of the choice, and a rating of how much to trust your gut
• Plan objection writer: Share a plan and your reasons. Get the strongest objections, a test for changing your mind and a decision call.
  • The best three arguments against you, in the opponent's words
  • The facts behind each, and which you can answer
  • The assumption of yours that is hit hardest
  • What would change your mind, and a call: go, change or wait
• Ruin risk stress tester: Check your plan for ruin risk. Paste it in and find the one outcome you could not recover from, plus a version with the downside capped.
  • A list of dependencies and things that could go wrong, with no odds guessed
  • The outcome you could not recover from, listed apart from ordinary losses
  • A capped-downside version of the plan, with the cost of the fix
  • What would make you act early
• Decision checklist runner: Run your checklist on a real decision. Each item gets a fact, a pass, fail or unknown, then a clear go, wait or no call.
  • Each item answered with evidence, marked pass, fail or unknown
  • Which stop items failed, if any
  • The call: go, wait or no, with a reason
  • A record of what you expect and what would show you wrong
• Checklist review planner: Share a decision record, the result and your checklist. Get a separate review of process and outcome, checklist changes and practice cases.
  • The call judged on process and outcome separately
  • Which items fired, which were missing, and what to add or cut
  • A version log of the changes
  • Three practice cases and a review schedule
• Lessons learned principle writer: Turn a mistake into a rule you can check. Give it what happened and your goal. You get the root cause, a rule and a check date.
  • A list of the problems without softening
  • A root cause about a choice or habit
  • A rule with the situation it applies to
  • Two past cases tested, a critic named and a check date
• Decision mistake collector: Log past decisions and outcomes. Get repeated causes, evidence limits and a prompt to interrupt the pattern next time.
  • Each case sorted: bad process, bad luck, lucky win or good call
  • A cause named for each real mistake
  • The causes that repeat, with a count
  • For each case, the question that would have caught it
• Decision checklist builder: Turn your mistake log into a decision checklist: up to twelve yes/no questions, each with a pass line, tested on two past cases.
  • A one-page checklist of twelve or fewer yes/no questions
  • A pass line and a place to find the answer for each
  • Stop items first, then cheap checks, then costly ones
  • A test of the list against two past cases
• Research paper and code note writer: Share a paper, repo or messy notes. Get claims, methods and evidence sorted into a clear research folder.
  • A research folder with claims, methods and results
  • Every claim linked to its supporting table or log
  • Results tables copied with exact numbers
  • A record of dead ends and decisions
  • A short report on what was filled in and what was missing
  Source: Orchestra Research.
• Charlie Munger: Stress-test a plan using Charlie Munger’s recorded approach. Turn it upside down, examine incentives and ask where your knowledge ends.
  • Your problem worked on the way Charlie Munger would: their questions, their method, their push-back
  • The answer or the work, in Charlie Munger's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step
• 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

Tasks you can use it for

• You face a decision that is hard to undo and want it checked the same way each time.
• You keep making the same kind of mistake and want a list that catches it.
• A past decision has played out and you want to learn from it fairly.

Limits to check before starting

• You face a small choice that is easy to undo: just decide, and check it later if it matters.
• You need money, legal or health advice: ask a qualified person, since this agent only gives a way to think.
• You do not yet have the facts: gather them first, because this agent will only send you back to find them.

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.

• Start a chat with the agent. In Claude Code, add it as a skill or save it in .claude/agents, then give it the decision.
• Answer its questions on facts it marks as unknown. It may ask twice.
• Make the call yourself. It saves the decision with a review date.
• On the review date, give it what happened. Say yes or no to each change it proposes for your checklist.
• The first time you have no checklist. After three decisions on file, it offers to build one.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/decision-checklist-keeper 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.