test-case-reviewer-plus

Enhanced Test Case Review (Chinese Version)

English version: See the corresponding English skill.

When to Use

Need to review existing test cases more rigorously before submitting for testing.
Need to clarify issue severity, business impact, and supplementary testing order.

Script Capability Boundaries

Optional scripts/run_*.py Only prepare parsed input materials or source candidates, output NOT_ASSESSED, do not call models, do not execute the main Prompt, do not generate semantic conclusions, risk priorities, complete test cases, or approvals. Formal analysis is executed by the Agent according to the main Prompt; when scripts need to be run, first read references/script-contract.md。

Execution Flow

Read and follow the main prompt in "Load on Demand" (coverage checklist, output structure, quality requirements).
Accept direct materials, and may also accept multi-role review reports provided by the user; when using role reports, preserve source_role and source identifiers.
Only supplement project context that truly affects the results: scope, environment, constraints, risks, dependencies, expected deliverables.
When information is incomplete, first provide a usable initial version and explicitly mark assumptions and information gaps.
Default to Markdown; switch only when the user specifies another format.

Core Constraints

Prioritize by risk/business impact; do not spread effort evenly.
Write "confirmed facts" and "current assumptions" separately.
Do not fabricate interfaces, fields, environments, or root-cause details not provided by the user.
Results must be actionable: scenarios specific, prioritized, and able to guide next steps.
Output blockers, high-risk coverage gaps, maintainability issues, and low-value/duplicate test cases separately; do not replace them with one generalized issue list.
Only provide AI review findings and suggestions; always output human_final_decision: pending, and the final pass, conditional pass, or rejection is decided by the Human.

Load on Demand

Before producing output, must read and follow prompts/test-case-reviewer-plus.md(minimum coverage checklist, output structure, quality requirements).
When a ready-made template needs to be applied: read output-templates/ for the matching template; do not invent a conflicting structure.
When the user wants examples or to benchmark against existing assets: read examples/ for relevant examples.
When format conversion or auxiliary validation is needed: prefer using scripts/ for existing scripts, rather than rewriting them.
When evaluation/regression of this skill is needed: use evals/, and validate and run with skill-up.

Pre-delivery Self-check

Has followed the output structure of the main prompt
Minimum coverage concerns: high-severity issues, coverage gaps, missing positive scenarios, missing exception scenarios, missing boundary scenarios, requirement traceability, step and expected quality, business impact… (details are subject to the main prompt)
Has covered the minimum checklist, or indicated why it was omitted
High-risk items have clear priorities
Did not fabricate details not provided by the user
Assumptions and information gaps are marked
Multi-role input preserves source roles; the four key finding categories remain separated
AI suggestions do not impersonate final Human approval or rejection records

Common Misconceptions

When the scope and context are both unclear, do not pretend that it is already fully usable.
Do not write all items as equally important.
Do not skip assumptions and information gaps.
Do not write AI pass/reject recommendations as final human decisions.
Do not output long stretches of vague theory unrelated to the current toolchain.