Hire an AI data analysis repeatability reviewer

Everything it can hand back

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

• An analysis inventory: what one row means, units, missing values, duplicates and exclusions
• A repeatable procedure: inputs, steps in order, settings, output files and what to do if a step fails
• A list of claims in your draft that the files do not support
• A list of gaps: what it could not see or run
• A hand-back summary and one next step

How it does the job well

• Turn every request into owned actions: who, what and by when.
• Work from the real calendar, notes and lists. Never invent meetings, people or dates.
• Put the most important decision or deadline first, then the details.
• Prepare drafts and proposals. Sending, booking and changing records waits for approval.

Mistakes it avoids

• Losing commitments buried in long notes.
• Mixing confirmed facts with guesses about other people's plans.
• Adding process when a short checklist would do.

What a good result looks like

An example, not your task: A colleague, reviewer or auditor asks how you got a number and you cannot show the steps.

• Analysis inventory. What each row means, units, missing values, duplicates and any rows left out.
• Reproduction procedure. Inputs, software, steps in order, settings, output files and what to check if a step fails.
• Unsupported claims list. Statements in your draft the files do not back up, with a safer wording for each.
• Gaps list. What the AI could not see or run, so nobody assumes it was checked.
• Hand-back summary. Files saved, MISSING lines, checks that could not run and one next step.

How the work is checked

• Only commands and outputs you actually saw are written as run. Described steps are marked as described.
• Every exclusion and every recalculation gap is visible in the inventory or the gaps list.
• Each step in the procedure names its input, its setting and its output.
• Every unsupported claim has a quote, a reason and a safer wording.
• 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.

Capabilities

• Spreadsheet checker. Inspects the data and any workbook for what a row means, units, missing values, duplicates and totals that reconcile. Covers only the file types and tools it can use. Hands the inventory to step 2.
• Process and SOP writer. Uses your own commands, workbook and software list, not step 1's file, and takes units and exclusions from the step 1 inventory. Writes the analysis as a procedure: inputs, software, order of steps, settings, output paths and failure checks. Original data stays untouched. Hands the procedure to step 3.
• Unsupported claim spotter. Reads your methods draft (your own text) and step 2's procedure for claimed checks or cause-and-effect statements that the files and run evidence do not show.

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 document-a-dataset-analysis-for-reproduction 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/document-a-dataset-analysis-for-reproduction/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 dataset file or files, and a data dictionary that says what each column means
• The commands, script or workbook you used for the analysis
• The software and its versions, such as Excel, R or Python
• The tables and charts the analysis produced
• Your draft methods note or report text (optional)

Helpful context (optional)

• A copy of the earlier report that used this analysis.
• A named reviewer who will try to repeat it.

Tools and access

• A spreadsheet engine or a code runner the AI can use: To recalculate a workbook or rerun a script, the AI needs the real tool. Without it, it reads and describes, and marks reruns as not done.

Functions and deliverables

• Data analysis repeatability reviewer: Review an analysis's files and methods. Document the actual steps, versions and outputs, and flag checks or findings that lack run evidence.
  • An analysis inventory: what one row means, units, missing values, duplicates and exclusions
  • A repeatable procedure: inputs, steps in order, settings, output files and what to do if a step fails
  • A list of claims in your draft that the files do not support
  • A list of gaps: what it could not see or run
  • A hand-back summary and one next step
• Spreadsheet checker: Check a spreadsheet for odd totals, broken formulas and messy data. Your original stays safe, and you get findings with proof and a new file.
  • A sheet inventory and scope
  • Findings with evidence locations
  • Data quality and formula issues
  • A new output file with your edits
  • Reconciliation results
  Source: Burhan Sebin.
• Process and SOP writer: Turn how your team works into a clear SOP. Give it a process name or old notes. You get a RACI table, a flow, steps and exceptions.
  • A step-by-step written procedure
  • A RACI table
  • A process flow
  • An exceptions and edge cases table
  • Metrics to track the process
  Source: Anthropic.
• Unsupported claim spotter: Find the claims in your draft that need a source before you publish. Each one is ranked by risk, with the kind of source to look for.
  • A summary of how many claims need support
  • A table of claims ranked by risk
  • The kind of source that would settle each
  • Internal conflicts in your draft
  Source: Pro Skill Packs.
• me.md builder: Answer a few questions, or let your AI use what it remembers. Get one private file that tells any AI who you are.
  • One me.md file with up to ten dated sections
  • A list of facts taken from memory, for you to check
  • A short note of what it removed for privacy
  • The two next steps, to save it and set it up

Tasks you can use it for

• A colleague, reviewer or auditor asks how you got a number and you cannot show the steps.
• You are handing an analysis to someone else to repeat next month or next year.
• Your methods note says "we checked for duplicates" and you want to know if the files prove it.
• You want to rerun last year's analysis on new data without guessing what you did.

Limits to check before starting

• You want formula errors checked before you share a workbook. Use the Spreadsheet results reviewer workflow.
• You want the top insights from a dataset. Use Dataset analysis and insight reviewer.
• You have not done the analysis yet. This documents one that is finished.
• You need to compare the methods of two research papers. Use Research paper disagreement 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 from the button on this page. It is a zip with HANDOFF.md and every skill file this workflow uses.
• Give the folder to Claude, ChatGPT or Gemini and say "Follow HANDOFF.md".
• Attach the data, the dictionary, your commands or workbook and your draft note. List your software versions.
• Read the inventory first. Correct any column the AI read wrongly.
• Try the procedure on a copy, or ask a colleague to. Note every step that does not work as written.
• Fix or cut each unsupported claim in your methods note before you share it.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/document-a-dataset-analysis-for-reproduction 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.