Nirmion
Help Find a tool

RESEARCH OPERATIONS LAB

Data Dictionary Builder

Create a field-level data dictionary with definition, type, values, null handling, source, owner, and classification from supplied metadata.

Browser processor readyRuns in this browser tabAccountable human review requiredUp to 500 rows using exactly 8 displayed input columns and 20,000 total characters.

Input

Build the data dictionary builder

Enter one complete row per line in the displayed column order. The exported CSV protects leading spreadsheet formula characters.

0 / 180

Name one bounded system, release, process, period, supplier, or programme. Avoid combining unrelated reviews.

0 / 180

Name the role responsible for validating the evidence and approving or escalating the final decision.

0 / 20,000

Follow the displayed pipe-separated order exactly. Do not put a pipe character inside a cell.

Required row orderField name | Business definition | Data type | Allowed values | Null handling | Source reference | Owner | Data classification

0 / 20,000

Optionally add approved evidence references, assumptions, exceptions, source dates, and follow-up context.

Review draft

Your reviewable output appears here

The data dictionary builder is generated only from the values you provide. No evidence, status, score, or approval is inferred.

  1. Load the fictional example to learn the exact input grammar.
  2. Replace it with one bounded set of approved facts and evidence references.
  3. Generate, inspect, download, then record the accountable decision elsewhere.

Product contract

Understand the data dictionary builder before use

This contract states the real input grammar, deterministic transformation, output format, and human-decision boundary for this route.

01

What it does

Create a field-level data dictionary with definition, type, values, null handling, source, owner, and classification from supplied metadata.

02

What you need

Complete privacy-minimized pipe-separated rows from an approved protocol, data plan, source schema, or research record, plus a bounded question, accountable owner, and method references.

03

What you receive

A separately named nirmion_tools file containing a formula-safe CSV register. Download resets current fields; the latest three small outputs remain available in this tab for re-download.

Practical workflow

A defensible data dictionary builder workflow

Use this route to create a field-level data dictionary with definition, type, values, null handling, source, owner, and classification from supplied metadata. It does not recruit people, collect responses, access repositories or datasets, clean data, test survey logic, perform statistical analysis, infer findings, or establish causality.

Before generation

Confirm the research question, protocol, consent and ethics status, population and sampling frame, source and period, privacy boundary, method, limitations, and accountable reviewer.

After generation

Validate the artifact against consent, protocol, source records, analysis plan, codebook, and quality checks; document exclusions and limitations; then store reviewed evidence and findings in authorized systems.

Likely edge cases

Each non-empty line must contain exactly 8 pipe-separated values. Empty cells, extra separators, more than 500 rows, or cells over 1,000 characters are rejected.

Decision boundary

A structured artifact is not a research finding

Do not treat this output as validated evidence, representative sampling, informed consent, cleaned data, statistical significance, causal proof, or a product decision. Qualified research, ethics, privacy, data, and domain reviewers must validate the method and interpretation.

Pre-use checklist

Review before operational use

  • Remove names, contact details, credentials, raw recordings, free-text secrets, health details, and unnecessary participant or customer identifiers; use approved pseudonymous references.
  • Confirm question, protocol, consent, ethics, sample, source, period, denominator, method, codebook, privacy controls, limitations, and accountable owner.
  • Check every row, statement, calculation, status, exception, mitigation, and follow-up action.
  • Store the reviewed artifact, analysis decisions, exclusions, limitations, evidence, approvals, and findings in the authorized research or data system.

Help and limitations

Data Dictionary Builder questions

Does this tool collect data, analyze a dataset, or produce findings?

No. It structures supplied metadata or performs one visible calculation. It does not recruit participants, host surveys, access datasets, clean records, test logic, estimate uncertainty, infer findings, or make a research decision.

Are inputs uploaded or retained?

No server request is used for generation. Entered text and the latest three generated outputs remain in this browser-tab instance until you refresh or close it. Do not enter information your policy forbids in a browser tool.

What happens after download?

The current form and result reset automatically. You can re-download the latest three generated text artifacts while this tab remains open, then validate and store the approved artifact in your official system.