test-plan-generatorlisted

# Test Plan Generator A generative skill: given a specific feature, change, user story, requirements document, API contract, or code diff, it produces a concrete, prioritized test plan document — not a findings table. Every other skill in this repo audits something that already exists and reports ✅/❌/⚠️/➖ rows; this skill instead writes new test cases that don't exist yet, so its output is a document deliverable (defined below), not a status table. **This skill is not `test-strategy-audit`.** `test-strategy-audit` asks whether a *project's* overall testing setup (pyramid balance, coverage tooling, CI gating, flakiness) is sound — it never produces concrete test cases for one feature. This skill is the opposite scope: one feature or change, concrete test cases, no opinion on the project's test infrastructure as a whole. If the user wants both, run them separately. ## Ground rules - **Ground every test case in real source material.** Use the actual acceptance criteria, user story, requirements doc, API contract/schema, or code diff/implementation being tested. Never invent generic filler cases like "test the happy path" or "test error handling" — every case must be concrete to the actual feature: specific inputs, specific expected outputs or state changes. If you don't have enough source material to make a case concrete, that's a gap to surface (see next rule), not a reason to write a vague one anyway. - **Don't silently guess at ambiguous or incomplete requiremen