Hire an AI budget, forecast and cash gap review

Everything it can hand back

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

• Each team's gap against budget with the reason, marked as the owner's or the agent's reading
• A refreshed full-year forecast and the gap against plan
• A 30, 60 and 90 day cash view against your floor
• Teams over for two or more months, with the pattern
• Six cash levers ranked by cash freed, only when the year is off track
• A quarterly survival score and what gives the plan room for error

How it does the job well

• Agree the question and the decision it supports before analysing.
• Check the data first: date range, grain, missing values, duplicates and definitions.
• Compare like with like: same period length, definitions and segments.
• Separate what the data shows from what it might mean, and give confidence levels.

Mistakes it avoids

• Recommending a cut because a month looked bad.
• Reading a timing shift (a bill paid early) as overspend.
• Editing the budget to make the gap disappear.
• Burying a cash problem under a list of small variances.
• Reporting a change without checking whether tracking or definitions changed.
• Treating correlation as cause.

What a good result looks like

An example, not your task: You run a small team with a budget by department or project and the monthly review keeps slipping.

• THE MONTH. Total budget, actual and gap, plus the full-year forecast set against the plan.
• CASH. The 30, 60 and 90 day cash view against your floor, with any dip under it shown first.
• MOVES BY TEAM. Each team's gap, the reason marked as the owner's or the agent's reading, and a note for the owner.
• REPEAT OFFENDERS. Teams over budget two or more months running, and the pattern.
• LEVERS. Six cash levers ranked by cash freed, each with a survival test, only when the year is off track.
• ROOM FOR ERROR. A quarterly survival score, where the plan breaks in a bad quarter, and what gives it room.

How the work is checked

• Every gap is budget minus actual from the files, not from memory.
• Each explanation says whether it came from an owner or is the agent's reading.
• No lever cuts something that cannot be restarted without a clear warning.
• The budget file was not changed.
• 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 Morgan Housel and Benjamin Graham would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Build the budget once. If none exists, turns past spend and targets into a phased budget by team, saved only when you approve it.
• Read the history. Opens the tracking rows and notes which teams have been over two months running.
• Explain the moves. Compares budget with actual by team and line, uses the owners' reasons and marks its own readings.
• Refresh the forecast. Recasts the full year from actuals so far and shows the gap against plan.
• Watch the cash. Builds the 30, 60 and 90 day view and puts any dip under your floor at the top.
• Rank the levers when needed. If the year is more than a tenth off or cash is at risk, ranks six levers by cash freed, each tested for survival.
• Check room for error quarterly. Scores whether the plan survives a bad quarter and what would give it room.
• Log and hand back. Saves the rows and levers, then stops for your yes.

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 monthly-budget-versus-actual-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/monthly-budget-versus-actual-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

• Last month's actual spend and income by team and line
• Your budget for the year by team and month, or past spend to build one from
• Bills due and what you are owed (advised)
• Your cash floor and the date the year must be back on plan (advised)
• One line from each team owner on what happened last month (optional)

Helpful context (optional)

• A list of contracts signed or hires started last month, so step changes are explained
• Last year's monthly spend, so seasonal patterns are known from the first run

Tools and access

• Claude Code: It can keep the budget, tracking and lever files between months and run on a schedule.
• Your accounting tool: Export actuals by department or class with the lines under each.
• Any good chat AI: Any chat AI can run the small model. Check the day 180 cash for the top lever by hand.
• Shopify: Its cohort and product reports give your repeat share and sell-through.
• Excel or Google Sheets: Holds the base model so you can change one lever and see the result.
• Any good chat AI: Stress-testing a plan in words and simple sums is text work.
• A spreadsheet: Lets you rerun the plan with a bad year and see cash by month.

Functions and deliverables

• Budget, forecast and cash gap review: Compare spending with budget, update the forecast and put cash risks first. Get proposed changes for review.
  • Each team's gap against budget with the reason, marked as the owner's or the agent's reading
  • A refreshed full-year forecast and the gap against plan
  • A 30, 60 and 90 day cash view against your floor
  • Teams over for two or more months, with the pattern
  • Six cash levers ranked by cash freed, only when the year is off track
  • A quarterly survival score and what gives the plan room for error
• Department budget planner: Build a department budget from past spending and targets. Get a monthly plan, variance reports and ideas to move money between teams.
  • Budget by department and line
  • Phased monthly plan
  • Variance report with causes
  • Reallocation suggestions
  Source: Burhan Sebin.
• Budget variance explainer: Find out why your actuals missed budget. Give it actual and budget numbers and get each gap split by cause, plus short notes for leadership.
  • A breakdown of each variance by cause
  • Volume, price and mix amounts
  • Short written commentary
  • A waterfall chart outline
  • A list of gaps worth action
  Source: Anthropic.
• Budget and forecast builder: Use past sales, costs, and plans to predict income, plan spending, and check budgets against real results.
  • A monthly or quarterly P&L forecast
  • A headcount and spend plan
  • Best, worst and base scenarios
  • What-if results
  • A budget against actuals tracker
  Source: GAJETOso.
• Cash forecast builder: Build a 30, 60 and 90 day cash forecast from your invoices, bills and fixed costs. See if you can make payroll, with risks named.
  • A 30, 60 and 90 day cash forecast with a low and high range for each
  • Up to five named risks, biggest dollar impact first
  • An Excel workbook with summary, detail and risk sheets
  • A short chat summary with the forecast table
  • A forecast page in the format you prefer
  Source: Anthropic.
• Cash flow improvement planner: Share stock, sales, costs and cash. Compare six changes by cash freed at 180 days and profit impact.
  • Cash at day 0, 30, 60 and 180 for each of six levers
  • The levers ranked by cash freed, with net income change shown apart
  • The best two or three to combine, and what to track for 90 days
  • A score out of 100 and how it was worked out
• Business plan risk buffer reviewer: Check your budget or business plan for room for error. Get a survival score out of 100, where it breaks in a bad year, and fixes.
  • A list of what must go right for the plan to work
  • Three believable ways it fails, each with a number or a date
  • The ruin point and a score out of 100 for how well the plan survives
  • Fixes sized to the worst case, and one rule against a panic move
• Metric definition writer: Share a metric, calculation and business use. Get a YAML definition, validation results and questions for the data owner.
  • A filled metric, dimension or entity YAML
  • A validation report
  • dbt-ready field names
  • Notes on edge cases
  Source: Nimrod Fisher.
• Morgan Housel: Review a plan’s room for error using Morgan Housel’s recorded ideas. Examine uncertainty, possible losses and what enough means in the case.
  • Your problem worked on the way Morgan Housel would: their questions, their method, their push-back
  • The answer or the work, in Morgan Housel's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step
• Benjamin Graham: Learn to separate price from value using Benjamin Graham’s recorded frameworks. Check the evidence, the downside and the room for error.
  • Your problem worked on the way Benjamin Graham would: their questions, their method, their push-back
  • The answer or the work, in Benjamin Graham'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 run a small team with a budget by department or project and the monthly review keeps slipping.
• One team is always a little over and you want to know whether it is timing or a pattern.
• Cash is tight and you want a cut list ranked by what it frees, not by what is easiest.

Limits to check before starting

• You only need a cash forecast: use the Cash forecast builder, or a simple sheet of bills and what you are owed.
• You have a one-off project with no budget: use a plain list of costs and a person to sign off each one.
• You want to close the books and work out profit: use the Monthly store profit 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.

• Start a chat with the agent. In Claude Code, add it as a skill or save it in .claude/agents, then ask it to run last month.
• Export last month's actuals by team from your accounting tool. Put the file and your budget in one folder, or attach them in the chat.
• Tell it your cash floor once. It saves it.
• Ask each team owner for one line on last month and paste the lines in. The explanations get much better.
• Read the moves and the levers. Approve what you agree with. Run it again on the 5th of next month.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/monthly-budget-versus-actual-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.