Test Data Requirement Analysis

Analyze entity, field, relationship, state, role, privacy, source, lifecycle, setup, and cleanup requirements before test design or data generation. Produce TDR-## requirements and blockers. Do not generate data, read production data, or copy the test-data-generation dataset plan.

When to Use

• Use it to determine which entities, fields, states, roles, and relationships a scenario needs.
• Use it to find schema, source, masking, isolation, lifecycle, cleanup, and environment prerequisite gaps.
• Use it when data material is incomplete but a constrained requirement list can be consumed by test design or data generation.

Do not use it to create records, call real data sources, copy production data, write full cases, or execute tests.

Output Format Options

• Use Markdown by default; when a table, CSV, or JSON is requested, preserve the same evidence, status, impact, owner, and validation fields.
• Do not present a structured format or static inventory as execution, pass, approval, or release evidence.

How to Use

1. Read this Skill's primary prompt and provide the objective, scope, material, environment, and available evidence.
2. Follow the prompt's input audit and output contract; deliver a bounded first pass when information is incomplete.
3. Retain source, evidence status, impact, owner role, close condition, and validation method for every finding.

Workflow

1. Read and follow prompts/test-data-requirement-analysis.md, beginning with the six-part input audit.
2. Map each scenario/test objective to required entities, fields, relationships, states, roles, sources, and environment.
3. Record valid/invalid/boundary/combination conditions, privacy/cleanup requirements, blockers, and validation in TDR-## entries.
4. Separate facts, preparation inference, open constraints, and Human/compliance decisions.
5. Deliver bounded requirements when information is missing and state why generation or execution cannot be promised.

Core Constraints

• Cover applicable entities/fields, referential integrity, valid/invalid/boundary/combination, states, roles, sources, masking, setup, cleanup, and isolation.
• Do not generate data, read/export real personal or production data, or connect to real data sources.
• Do not invent fields, relationships, quantities, privacy rules, retention, permissions, or cleanup results.
• TDR-## is a preparation requirement/gap, not proof that data exists, is available, compliant, or passed testing.
• State what later test design or data generation needs; do not replace generation or Human/compliance approval.

Reference Files

• Always read prompts/test-data-requirement-analysis.md before producing an analysis.
• For regression, read evals/eval.yaml and matching cases; configuration does not prove data is ready.
• For trigger checks, use evals/trigger-prompts.csv and evals/local-rules.json; missing selection trace is BLOCKED.

Best Practices

• Prioritize high-impact gaps with a verifiable next action, using the smallest useful experiment or evidence request.
• Separate facts, evidence-backed inferences, recommendations, and Human decisions; never upgrade an assumption into a conclusion.

Pre-delivery Check

• Recorded known facts, missing information, conflicts, stale information, out-of-scope items, and assumptions.
• Each TDR-## has scenario/goal, entities/fields, relationships/states/roles, constraints, source, and evidence.
• Valid/invalid/boundary/combination, masking, setup, isolation, and cleanup requirements are separate where applicable.
• Blockers have an owner role, smallest evidence action, and validation method.
• No records were generated, production data read, or requirements presented as existing data/passed testing.

Common Pitfalls

• Listing values without explaining why they are needed, related, and cleaned up.
• Treating schema presence as data availability or a masking statement as compliance evidence.
• Mixing data-preparation requirement analysis with dataset generation.
• Ignoring roles, states, referential integrity, and environment isolation while checking only field format.