test-analyst

Converts acceptance criteria and walked scenarios into executable, traceable test cases.

1|Updated Jun 30, 2026
One-click install
npx skills add https://github.com/Lia-Creative/lia-plugins --skill test-analyst-lia-creative
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: test-analyst
Source: https://github.com/Lia-Creative/lia-plugins/tree/main/lia-tools/skills/test-analyst
Command: npx skills add https://github.com/Lia-Creative/lia-plugins --skill test-analyst-lia-creative

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Frozen acceptance criteria and walked scenarios describe what must be true, but nobody can run them directly. This Skill turns them into a concrete case set — one case per criterion plus the implied seams — each with preconditions, real inputs, steps, and a decidable expected result, so a separate tester can execute without asking questions. ## Core Features & Use Cases - Criterion-to-case traceability: Every case carries an index and cites the acceptance criterion it covers, so coverage gaps and bug reports trace both ways. - Seam coverage beyond the criteria: Adds cases for the state after the action, the second run over the same data, and interrupted paths that criteria imply but never state. - Write/run separation: Cases are filed as a PR to the testing lead and never executed by the writer, so the run can genuinely fail the writing. - Use Case: After a story's acceptance criteria are frozen and its flows walked, invoke this Skill to produce the runnable case set the QA stage will execute against the real build. ## Quick Start Ask the agent to turn this story's frozen acceptance criteria and walked scenarios into the executable test case set, one case per criterion plus the seams.

Frequently Asked Questions about test-analyst

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

FAQPage Schema
How do I turn acceptance criteria into test cases?▼

Write at least one case per acceptance criterion, each with an index, preconditions, concrete steps with real input values, and one observable expected result. Then add cases for the seams the criteria imply, such as the state after the action and the second run over the same data.

What makes a test case executable by someone else?▼

An executable case uses concrete inputs rather than placeholders, states preconditions the runner can create, and defines a decidable expected result. If the verdict depends on the runner's judgement, it belongs in the plan's open questions instead.

Should test case writers also run the cases they write?▼

No. Writing and running are split so the run can fail the writing — a case set executed by its own author is graded by the person who chose the questions. This Skill files cases as a PR to the testing lead and never runs them.

What happens when a scenario walk is missing for a criterion?▼

A missing walk is commissioned back through the testing lead to discovery's scenario-builder, not improvised during case writing. A QA-invented walk would compete with the discovery artefact and produce two documents describing the same journey differently.

Does this replace automated test code?▼

No. Automated tests belong to the build stage and are written with the feature. This Skill produces the manual case set the QA stage executes against the real build, distinct from code-level test automation.