cmd-speckit-analyze

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

/speckit.analyze

Agent skill wrapper for the Claude command

/speckit.analyze

When the original command text references

{{INPUT}}

$1

, or named arguments, map them from the user's current request.

用于Claude命令

/speckit.analyze

的Agent技能封装器。

当原始命令文本引用

{{INPUT}}

、

$1

或命名参数时，从用户当前请求中映射对应内容。

Command Instructions

命令说明

Read-only consistency check across all specification artifacts for a feature. No files are modified.

对某个功能的所有规范工件进行只读一致性检查。不会修改任何文件。

Steps

步骤

Locate artifacts:
- If input names a feature or directory, use it.
- Otherwise check
```
.specify/specs/
```
  and ask if multiple features exist.
- Load all available:
```
spec.md
```
  ,
```
plan.md
```
  ,
```
tasks.md
```
  ,
```
data-model.md
```
  ,
```
contracts/*.md
```
  ,
```
research.md
```
  .
- Load
```
.specify/memory/constitution.md
```
  .
Run six consistency checks. Record each finding with: category, severity (CRITICAL / HIGH / MEDIUM / LOW), location, description.
- Constitution conflicts (auto-CRITICAL): Any plan or task that violates a MUST or SHOULD NOT principle.
- Coverage gaps: Requirements with no corresponding tasks; tasks traceable to no requirement.
- Inconsistency: Terminology drift, conflicting tech choices, ordering contradictions between artifacts.
- Ambiguity: Vague terms (
```
fast
```
  ,
```
scalable
```
  ,
```
secure
```
  ) without measurable criteria; unresolved
```
[NEEDS CLARIFICATION]
```
  markers.
- Underspecification: Incomplete acceptance criteria, undefined component references, missing error conditions in contracts.
- Duplication: Near-identical requirements or tasks appearing in multiple places.

Produce analysis report in the conversation (do not write to any file):

## Consistency Analysis: <Feature Name>

### Findings (sorted CRITICAL → HIGH → MEDIUM → LOW, max 50 rows)
| # | Severity | Category | Location | Finding |

### Coverage Matrix
Each spec.md requirement → mapped tasks (or NONE)
Each tasks.md task → source requirement (or NONE)

### Metrics
- Requirement coverage: X%
- Constitution violations: N
- Ambiguous items: N
- Duplicate items: N

### Recommended Next Actions

Offer remediation: If findings exist, ask: "Would you like me to fix any of these?" Apply only user-approved edits.

Next step: Address any CRITICAL findings, then run

/speckit.tasks

定位工件:
- 如果输入指定了功能或目录，则使用该输入。
- 否则检查
```
.specify/specs/
```
  目录，并询问是否存在多个功能。
- 加载所有可用文件：
```
spec.md
```
  、
```
plan.md
```
  、
```
tasks.md
```
  、
```
data-model.md
```
  、
```
contracts/*.md
```
  、
```
research.md
```
  。
- 加载
```
.specify/memory/constitution.md
```
  。
执行六项一致性检查。记录每个检查结果，包含：类别、严重程度（CRITICAL / HIGH / MEDIUM / LOW）、位置、描述。
- 章程冲突（自动标记为CRITICAL）：任何违反MUST或SHOULD NOT原则的计划或任务。
- 覆盖缺口：无对应任务的需求；无法追溯到需求的任务。
- 不一致性：术语偏差、技术选择冲突、工件间的顺序矛盾。
- 模糊性：无衡量标准的模糊术语（如
```
fast
```
  、
```
scalable
```
  、
```
secure
```
  ）；未解决的
```
[NEEDS CLARIFICATION]
```
  标记。
- 规范不足：不完整的验收标准、未定义的组件引用、合约中缺失的错误条件。
- 重复内容：在多个位置出现的近乎相同的需求或任务。

在对话中生成分析报告（请勿写入任何文件）：

## 一致性分析：<功能名称>

### 检查结果（按CRITICAL → HIGH → MEDIUM → LOW排序，最多50行）
| 序号 | 严重程度 | 类别 | 位置 | 结果描述 |

### 覆盖矩阵
每个spec.md中的需求 → 对应的任务（或NONE）
每个tasks.md中的任务 → 来源需求（或NONE）

### 指标
- 需求覆盖率：X%
- 章程违规次数：N
- 模糊项数量：N
- 重复项数量：N

### 建议后续操作

提供修复建议：如果存在检查结果，询问：“是否需要我修复其中的问题？”仅执行用户批准的修改。

下一步：处理所有CRITICAL级别的检查结果，然后运行

/speckit.tasks

。