Data Dictionary

Skill Name: data-dictionary

Background & Use Cases

When requirements mention business objects like "accounts," "orders," and "transaction records," the development team and architects need to know what properties these objects have, their relationships, and where data comes from and goes to. Without unified data definitions, frontend, backend, and database teams may each use different field names, causing massive rework during integration. This skill helps you organize these business objects into a structured data dictionary.

Input & Output

You need to provide: Scenario descriptions and data entity list from the requirements specification, plus the external system inventory.

You will receive: A data dictionary document containing: ER diagram (entity relationship diagram), field definitions and specifications for each data entity, and data flow diagrams (how data moves between systems).

Process Overview

The Agent extracts all data entities from the requirements specification, then goes through each: what fields does this entity have, what type is each field, what constraints apply, what are the relationships between entities, which system does the data come from and flow to. The final output includes ER diagrams, field specifications, and data flow diagrams.

flowchart TD s0["Preparation & Scope Definition"] s1["Data Store Inventory"] s2["Data Flow Diagram Creation"] s3["Field Decomposition"] s4["Field Standardization"] s5["Data Structure & Relationships"] s6["Encoding & Value Domains"] s7["Document Integration & Full Review"] s8["Review & Publication"] s0 --> s1 s1 --> s2 s2 --> s3 s3 --> s4 s4 --> s5 s5 --> s6 s6 --> s7 s7 --> s8