Write a data dictionary for a table or dataset
Produces clear column-level documentation with definitions, types, allowed values, grain, and known caveats so others can use the data safely.
The prompt
Write a data dictionary for a table or dataset so other people can use it without asking me.
Table/dataset name and purpose: [e.g. "analytics.fct_orders - one row per order line"]
Schema: [PASTE column names and data types, or a CREATE TABLE statement]
Sample rows: [PASTE 5-20 rows, with sensitive values masked]
How it is built: [source systems, refresh schedule, key transformations - or "unknown"]
Who will use it: [analysts, business users in a BI tool, data scientists]
Produce:
1. Overview: what one row represents (grain), primary key, refresh frequency, owner [ADD], and the main use cases.
2. Column table: Column | Type | Plain-English definition | Example value | Allowed values or range | Nullable and what null means | Notes.
3. Relationships: how it joins to other tables, on which keys, and the join cardinality (one-to-one, one-to-many).
4. Gotchas: fields that are commonly misused, look similar but differ, contain test or internal records, or change meaning over time.
5. Example questions this table can and cannot answer.
Rules:
- Infer definitions from names and sample values, but mark anything you are guessing as [CONFIRM] rather than stating it as fact.
- Flag columns that look like personal or sensitive data.
- Keep definitions short and free of jargon so a business user can read them.
Highlighted [brackets] are blanks to replace with your own details.
Fill in the blanks and use this promptMore Data & Analysis prompts
What to fill in
- e.g. "analytics.fct_orders - one row per order line"
- PASTE 5-20 rows, with sensitive values masked
- analysts, business users in a BI tool, data scientists
- ADD
- CONFIRM
How this prompt is built
This prompt follows the RTF structure (Role, Task, Format). RTF tells the model who to be, what to do and what shape the answer should take. It is the quickest structure that still gives noticeably better results than a bare request.
Related prompts
- Write up statistical results in proper reporting styleTurns software output into a correctly formatted results section with effect sizes, confidence intervals, and matching tables.
- Write commentary for a dashboard or KPI reportWrites the 'so what' narrative executives need alongside a dashboard — for sales, marketing, HR, finance, ops, or support.
- Write a SQL query from plain EnglishTurns a business question into SQL, with validation checks and an explanation.
- Write a monthly IT and security report for leadershipTurns ticket, uptime, and security metrics into a short report showing trends, risks, and decisions needed.