Requirements Traceability TraceMap

将 PRD、规格、验收标准、测试用例与多仓库代码变更映射为需求-测试-代码追溯矩阵，识别已覆盖、部分覆盖、未覆盖、冗余与阻塞需求，输出缺陷清单、覆盖率缺口、发布验收建议，并覆盖 LLM、Agent 与 RAG 工作流的验证。

start-fish

热度 16.02

最近更新 2026-06-17 11:30:18

分类：开发依赖：读写文件应用场景：代码审查 / 报告撰写 / 工作流

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add tracemap-ai

Metadata

name: tracemap-ai
description: Requirements Traceability and Test Case Validation skill for PRD analysis, requirement coverage, coverage gap detection, multi-repo testing, acceptance testing, traceability matrix, LLM testing, Agentic AI, and RAG testing.
triggers: ["requirements-traceability","test-case-validation","requirement-coverage","coverage-gap","acceptance-testing","multi-repo-testing","qa-report","traceability-matrix","defect-analysis","prd-analysis","spec-validation","agentic-ai","rag-testing","llm-testing"]
role: specialist
scope: testing
output-format: report

TraceMap AI

TraceMap AI verifies whether requirements, test cases, and code changes truly line up. It converts PRDs, specs, acceptance criteria, test cases, and multi-repo code changes into a requirement-test-code traceability matrix.

It works for product features, backend services, frontend flows, mobile apps, APIs, data pipelines, AI workflows, LLM integrations, agents, RAG systems, and cross-repository releases.

When To Use

Use this skill when you need to:

Validate whether a PRD or spec is fully covered by test cases and code changes.
Review test cases before development, test execution, or release acceptance.
Analyze multi-repo changes across frontend, backend, services, configuration, prompts, data pipelines, and integrations.
Build traceability evidence for QA sign-off, release governance, or product acceptance.
Identify uncovered requirements, weak acceptance criteria, missing evidence, and high-impact defects.

Core Workflow

Collect inputs: PRD, spec, acceptance criteria, test cases, repositories, Git Diff, configuration, prompt/model changes, and release scope.
Extract requirements: split product and technical requirements into atomic, testable acceptance points.
Map implementation: connect each requirement to code files, services, APIs, configs, prompts, models, and data flows.
Map tests: connect each requirement to test cases, fixtures, expected evidence, and execution results.
Analyze coverage: classify covered, partially covered, uncovered, redundant, blocked, and untestable requirements.
Analyze defects: identify severity, priority, reproduction steps, business impact, and owner suggestions.
Generate report: produce a traceability matrix, gap summary, defect list, and release acceptance recommendation.

Reference Guides

Topic	Reference	Load When
Traceability validation	`references/traceability-validation-playbook.md`	Always use for coverage categories, mapping dimensions, and matrix template

Required Output

Produce a Markdown report with:

Executive summary and release acceptance recommendation.
Requirement-test-code traceability matrix.
Coverage summary: covered, partial, uncovered, redundant, blocked.
Defect list with severity, priority, reproduction steps, owner, and fix recommendation.
AI workflow validation matrix when applicable.
Evidence checklist and open questions.

AI Workflow Validation Matrix

## AI Workflow Validation Matrix

| Requirement ID | Scenario | AI/Model/Agent Link | Test Case | Code Location | Evidence | Coverage Result | Gap |
| --- | --- | --- | --- | --- | --- | --- | --- |

Quality Bar

Do not mark a requirement as covered unless there is clear test and code evidence.
Distinguish missing implementation from missing tests.
Every defect must include impact and reproduction path.
Every AI workflow must include fallback, observability, and safety checks.
If evidence is incomplete, state what needs to be provided.