Document a data analysis with AI
Turn source records, methodological choices and results into a handover that explains both the business decision and the steps behind the numbers.
Document a data analysis with AI before the reasoning disappears into a notebook nobody else can follow. A sales total can be reproduced only if the next analyst knows which cancellations were excluded and which date column determined the reporting period. The analysis-documentation skill organises those choices beside the findings.
Use the material from the actual work. A business reader needs to understand what the conclusion supports, while a technical reviewer needs the source tables, method and environment. A short opening and a detailed technical section can serve both.
What it does
- State the question and the decision it informs.
- Record source periods, key columns and known quality issues.
- Explain important choices and the alternatives considered.
- Document limitations that change how a result can be used.
- Write a reproducible handoff for someone returning to the analysis later.
Install and use
- Give the agent authorised source descriptions, method notes and existing results. Creating the documentation needs no live database connection.
- Ask, “Prepare this sales analysis for handover. Show the cancellation rule, the reporting date and anything the original analyst still needs to confirm.”
- Check one number against the documented route and correct any unsupported claim before circulating the document.
FAQ
Source
Made by Nimrod Fisher. Original skill: analysis-documentation, published on GitHub under the MIT licence.
View on GitHub