From omni
分析指定模块目录的单一职责原则遵守情况,生成三维度得分与违规报告的JSON。适用于检查模块内聚性、检测职责混杂问题。
How this skill is triggered — by the user, by Claude, or both
Slash command
/omni:ai-friendly-arch-guard-module-single-responsibilityThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
分析指定模块目录的单一职责原则遵守情况,输出标准化JSON格式报告。
分析指定模块目录的单一职责原则遵守情况,输出标准化JSON格式报告。
必须使用subagent执行此技能,通过Agent tool调用,subagent_type为"general-purpose"。
| 参数 | 类型 | 必填 | 说明 |
|---|---|---|---|
module_path | string | ✅ | 待分析的模块目录路径(绝对路径或相对路径) |
scan_mode | enum | ❌ | 扫描模式:full(全量)或 increment(增量),默认 increment |
output_path | string | ❌ | 分析结果的保存路径。由编排 skill 调用时传入具体路径(如 state/step02-analyze-modules/{module_name}.json);单独调用时默认保存到 .claude/skills/ai-friendly-arch-guard-module-single-responsibility/output/analysis_result.json |
输出严格遵循 schema/output-schema.json 定义的 JSON Schema(完整字段定义见该文件)。
关键字段摘要:
| 字段路径 | 类型 | 说明 |
|---|---|---|
skill_id | string | 固定值 |
rule_version | string | 当前 v1.3 |
module_path | string | 模块目录相对路径(供编排层读取) |
scan_mode | enum | increment | full |
execute_status | enum | success | failed |
metric_result.total_score | number(0-100) | 综合得分 |
metric_result.confidence | number(0-1) | 置信度 |
metric_result.score_detail | object | 三维度得分明细 |
metric_result.confidence_detail | object | 三维度置��度明细 |
violation_info.total_count | integer | 违规总数 |
violation_info.level_summary | object | P0/P1/P2 分级统计 |
violation_info.list | array | 违规详情列表 |
violation_info.exempt_list | array | 豁免列表 |
statistic_info | object | 扫描单元统计(目录/模块/文件) |
| 等级 | 分数区间 | 含义 |
|---|---|---|
| S 级 | 90-100 | AI 原生友好,可全自动开发 |
| A 级 | 75-89 | AI 友好优良,适合 AI 辅助开发 |
| B 级 | 60-74 | 基础可用,需适度优化 |
| C 级 | 40-59 | AI 开发困难,需重构 |
| D 级 | 0-39 | 不适合 AI 开发 |
npx claudepluginhub zte-aicloud/co-omnispec --plugin omniOrchestrates multi-module SRP (Single Responsibility Principle) compliance analysis across a codebase. Supports full-scan and git-diff-based incremental mode for CI gating. Identifies modules, runs concurrent SRP checks, then aggregates results into a summary report.
Analyzes codebase architecture via multi-agent specialists on structure, coupling, integration, error handling, security; verifies findings, reports strengths and flaws with evidence.
Maps module dependencies, checks layering integrity, and flags structural decay across a codebase. Includes onboarding mode for codebase tours.