sdlc-testcase

Generates human-readable test case catalogs from SRS scenarios and tech design test strategies.

Updated Jun 1, 2026
One-click install
npx skills add https://github.com/juartaurus98/codebase --skill sdlc-testcase-juartaurus98
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sdlc-testcase
Source: https://github.com/juartaurus98/codebase/tree/main/.claude/skills/sdlc-testcase
Command: npx skills add https://github.com/juartaurus98/codebase --skill sdlc-testcase-juartaurus98

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It bridges the gap between requirements and test automation by producing a reviewable, human-readable test case catalog before any test code exists, ensuring coverage decisions happen at design time rather than during implementation. ## Core Features & Use Cases - Test Case Catalog Generation: Derives happy-path, edge, negative, and non-functional test cases from .feature scenarios and the tech design's test strategy section, with checkable expected results. - AI Evaluation Cases: Authors eval cases in JSONL format (golden, adversarial, regression, prod-sample) for LLM features, with structural expectations and judge rubric minimums. - Explicit Non-Automation Tracking: Documents what will not be automated with reasons, owners, and expiry dates so coverage gaps are visible rather than implied. - Use Case: After finalizing SRS scenarios and a tech design, invoke /gen-testcase with a UC-ID to produce a TC catalog that /unittest and /run-test later partition and automate. ## Quick Start Run /gen-testcase with a use case ID such as UC-LOGIN-01 to generate the test case catalog from the existing scenarios and tech design.

Frequently Asked Questions about sdlc-testcase

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate test cases from requirements before writing test code?▼

Invoke /gen-testcase with a UC-ID or FEAT-ID. It reads the .feature scenarios and the tech design's test strategy section, then writes a catalog with columns for level, preconditions, steps, test data, expected results, and oracle.

What is the difference between a test case catalog and test code?▼

A test case catalog is a human-readable specification of preconditions, steps, and expected results that people can review and execute by hand. This skill produces only the catalog; downstream phases like /unittest and /run-test write and run the actual test code.

How do I write evaluation cases for LLM features?▼

For scenarios marked Level: eval, the skill authors cases.jsonl files with golden, adversarial, regression, and prod-sample origins. Each case carries structural expectations like json_schema or must_contain, plus optional judge rubric minimums.

Can the test case generator change a scenario's test level or oracle?▼

No. The Level and Oracle assigned in the tech design's section 9 are authoritative. Disagreement is reported as a finding routed to the techdoc phase rather than silently changed during test authoring.

What happens when a test case cannot be automated?▼

It is listed in a Not automated section with a reason, such as physical devices or third-party sandboxes. Each entry needs a matching waiver row with an owner and expiry, or traceability reporting flags it as a missing test.