Speckit Analyze Zh

This is a read-only analysis tool for Chinese project documentation that runs after you've generated your spec, plan, and tasks files. It cross-checks all three artifacts for inconsistencies, duplicate requirements, vague acceptance criteria, and gaps in task coverage. The skill also validates against a project constitution file to catch principle violations early. You get a structured markdown report with severity ratings, coverage metrics, and remediation suggestions, but nothing gets modified automatically. Use this before implementation when you want to catch planning errors like unmapped requirements, conflicting tech stack choices, or tasks that reference undefined components. It's designed to stay under token limits by loading only essential context and capping findings at 50 items.

skills.sh installs: 308
GitHub stars: 10
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add forztf/open-skilled-sdd --skill speckit-analyze-zh --agent claude-code

Installs into .claude/skills of the current project.

Files

用户输入

$ARGUMENTS

您必须在继续之前考虑用户输入（如果不为空）。

目标

在实现之前，识别三个核心工件（spec.md、plan.md、tasks.md）之间的不一致、重复、歧义和未充分说明的项目。此命令必须仅在 speckit-tasks 成功生成完整的 tasks.md 后运行。

操作约束

严格只读：不要修改任何文件。输出结构化分析报告。提供可选的补救计划（用户必须明确批准后才能手动调用任何后续编辑命令）。

宪章权威性：项目宪章（.specify/memory/constitution.md）在此分析范围内是不可协商的。宪章冲突自动为关键级别，需要调整规格、计划或任务——而不是稀释、重新解释或默默忽略原则。如果原则本身需要更改，必须在 speckit-analyze 之外的单独、明确的宪章更新中进行。