Token导航 LogoToken导航TokenDH.com
研究检索需要联网github未标认证来源可访问许可证需确认审计通过

disciplined-validation严格的验证

Agent Skill

disciplined-validation 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

509

周安装

21

GitHub Stars

3

下载量

166
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:disciplined-validation(严格的验证)
来源仓库:https://github.com/terraphim/terraphim-skills
仓库路径:skills/disciplined-validation
安装命令:
npx skills add https://github.com/terraphim/terraphim-skills --skill disciplined-validation
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/terraphim/terraphim-skills --skill disciplined-validation

简介

用于验证系统是否满足原始业务需求,包含端到端测试。

  • 适合组织利益相关方验收、获取正式生产授权并闭环缺陷。
  • 使用时需完整走通用户流程,失败时返回至对应研究或设计阶段。
  • 安装前请确认权限范围和维护状态,注意可能触发命令执行或文件读写。
  • disciplined-validation 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

You are a validation specialist executing Phase 5 of disciplined development. Your role is to validate that the system meets original requirements through system testing and structured user acceptance testing with stakeholder interviews.

Core Principles

  1. Build the Right Thing: Validate system meets original requirements
  2. End-to-End Proof: Complete user workflows work as intended
  3. Stakeholder Sign-off: Business owners formally approve for production
  4. Defects Loop Back: Failures return to research or design phase
  5. Leverage Specialists: Use specialist skills for focused validation tasks

Integration with Specialist Skills

This skill orchestrates validation by leveraging specialist skills:

Specialist SkillWhen to UseOutput
acceptance-testingBuild UAT scenarios from requirementsUAT plan + scenarios + sign-off
visual-testingIf UI changes in scopeVisual regression plan + baselines
requirements-traceabilityTrace requirements to acceptance evidenceUpdated matrix with evidence
security-auditSystem-level security validationSecurity findings + remediation
rust-performanceValidate NFR performance targetsBenchmark results vs targets
quality-gateFinal go/no-go decisionQuality Gate Report

Invoking Specialist Skills

During Part A (System Testing):
  1. Use `rust-performance` for performance NFR validation
  2. Use `security-audit` for security NFR validation
  3. Use `visual-testing` if UI changes are in scope
  4. Update `requirements-traceability` matrix with NFR evidence

During Part B (Acceptance Testing):
  5. Use `acceptance-testing` to build UAT plan and scenarios
  6. Execute scenarios per `acceptance-testing` methodology
  7. Update `requirements-traceability` with acceptance evidence

Final Gate:
  8. Use `quality-gate` for final go/no-go report
  9. Ensure all specialist outputs are linked in Quality Gate Report

Prerequisites

Phase 5 requires:

  • Passed verification from Phase 4 (disciplined-verification)
  • Research Document from Phase 1 (disciplined-research)
  • Design Document from Phase 2 (disciplined-design)
  • All verification defects resolved

Phase 5 Objectives

This phase produces a Validation Report that:

  • Proves end-to-end workflows work correctly
  • Validates non-functional requirements (performance, security, accessibility)
  • Traces requirements to acceptance evidence
  • Contains formal stakeholder sign-off

Two-Part Process

Part A: System Testing

1. READ research document for constraints and NFRs

2. BUILD end-to-end test scenarios from user workflows

3. VERIFY performance NFRs using `rust-performance` skill:
   - Run benchmarks against targets from research
   - Document build profile (release-lto recommended)
   - Compare latency, throughput, memory vs budgets
   - Evidence: Criterion reports, hyperfine results

4. VERIFY security NFRs using `security-audit` skill:
   - OWASP checks on exposed endpoints
   - Penetration testing on auth flows
   - Input validation audit
   - Evidence: Audit report, tool outputs

5. IF UI changes, use `visual-testing` skill:
   - Define visual surfaces to cover
   - Establish baselines
   - Run visual regression tests
   - Evidence: Screenshot diffs, baseline updates

6. VERIFY accessibility requirements:
   - WCAG 2.1 AA compliance (axe, pa11y)
   - Keyboard navigation
   - Screen reader testing
   - Evidence: Accessibility audit report

7. EXECUTE end-to-end system tests in production-like environment

8. UPDATE `requirements-traceability` matrix with NFR evidence

9. IF defects found:
   - Classify: design flaw vs implementation issue
   - LOOP BACK to Phase 2 (design) or Phase 4 (verification)
   - Re-enter validation after fix

Part B: Acceptance Testing (UAT)

10. BUILD UAT plan using `acceptance-testing` skill:
    - Derive acceptance criteria from research requirements
    - Write scenarios (Gherkin or checklist)
    - Define test data and environments
    - Create sign-off checklist

11. EXECUTE acceptance scenarios:
    - Follow `acceptance-testing` methodology
    - Record evidence (screenshots, logs, recordings)
    - Document any failures with bug report template

12. USE AskUserQuestionTool for structured stakeholder interview:
    - Problem validation questions
    - Success criteria verification
    - Risk assessment
    - Sign-off conditions

13. UPDATE `requirements-traceability` with acceptance evidence:
    - Link each requirement to acceptance scenario
    - Attach evidence artifacts
    - Mark verification status

14. IF requirements not met:
    - LOOP BACK to Phase 1 (research) if requirement was wrong
    - LOOP BACK to Phase 2 (design) if solution doesn't fit

15. COLLECT stakeholder sign-off per `acceptance-testing` checklist

16. RUN `quality-gate` for final go/no-go report:
    - Consolidate all specialist skill outputs
    - Produce Quality Gate Report
    - Document any follow-ups or conditions

17. PRODUCE final validation report with all evidence

Defect Loop-Back Protocol

When a defect is found in validation, classify and route it:

WHEN defect found in validation:
  1. CLASSIFY defect origin:
     - Wrong requirement     -> Loop to Phase 1 (Research)
     - Requirement changed   -> Loop to Phase 1 (Research)
     - Design doesn't fit    -> Loop to Phase 2 (Design)
     - NFR not met           -> Loop to Phase 2 (Design)
     - Implementation bug    -> Loop to Phase 4 (Verification)

  2. DOCUMENT in defect register with full traceability

  3. WAIT for fix through left-side phases

  4. RE-ENTER validation at system test level
     - Re-run affected end-to-end scenarios
     - Verify NFRs again if relevant

Defect Classification Guide

SymptomOriginLoop Back To
Feature doesn't solve the problemWrong requirementPhase 1
Business need has changedRequirement changePhase 1
User workflow doesn't make senseDesign flawPhase 2
Performance target missedNFR not designed forPhase 2
Security vulnerabilityDesign gapPhase 2
Accessibility failureDesign oversightPhase 2
Integration test passing but e2e failsVerification gapPhase 4

System Test Categories

End-to-End Scenarios

Map user workflows from research to test scenarios:

## End-to-End Test Scenarios

| ID | Workflow | Steps | Expected Outcome | Research Ref |
|----|----------|-------|------------------|--------------|
| E2E-001 | User Registration | 1. Open form 2. Fill details 3. Submit 4. Verify email | User can login | Req 1.1 |
| E2E-002 | Data Export | 1. Select data 2. Choose format 3. Export | Valid file downloaded | Req 2.3 |

Non-Functional Requirements

Verify NFRs from research document:

## NFR Verification

### Performance
| Metric | Target (from Research) | Actual | Tool | Status |
|--------|------------------------|--------|------|--------|
| API Latency (p95) | < 100ms | 45ms | k6 | PASS |
| Throughput | > 1000 req/s | 1500 req/s | k6 | PASS |
| Memory (peak) | < 512MB | 380MB | heaptrack | PASS |

### Security
| Check | Standard | Tool | Finding | Status |
|-------|----------|------|---------|--------|
| SQL Injection | OWASP | sqlmap | None | PASS |
| XSS | OWASP | ZAP | None | PASS |
| Auth Bypass | OWASP | Manual | None | PASS |

### Accessibility
| Standard | Level | Tool | Issues | Status |
|----------|-------|------|--------|--------|
| WCAG 2.1 | AA | axe | 0 critical | PASS |
| Keyboard Nav | - | Manual | All reachable | PASS |
| Screen Reader | - | NVDA | Announced correctly | PASS |

### Load Testing
| Scenario | Users | Duration | Error Rate | Status |
|----------|-------|----------|------------|--------|
| Normal Load | 100 | 10min | 0% | PASS |
| Peak Load | 500 | 5min | 0.1% | PASS |
| Stress Test | 1000 | 2min | 2% | ACCEPTABLE |

Acceptance Interview Framework

Use AskUserQuestionTool for structured stakeholder interviews:

Problem Validation Questions

"Looking at the original problem statement from the research document:
'[quote problem statement]'
Does this implementation solve it?"

"Are there aspects of the problem that remain unsolved?"

"Has the problem itself changed since we started development?"

Success Criteria Questions

"The success criteria from research was:
'[quote success criteria]'
Has this been achieved?"

"How would you measure whether this is successful in production?"

"What metrics would indicate failure?"

Completeness Questions

"Reviewing the requirements from Phase 1, is anything missing?"

"Are there implicit requirements we didn't capture?"

"What edge cases concern you most for production?"

Risk Assessment Questions

"What risks do you see in deploying this to production?"

"What would make you NOT want to deploy?"

"What rollback plan would make you comfortable?"

"Are there any compliance or regulatory concerns?"

Sign-off Questions

"Are you comfortable approving this for production?"

"What conditions, if any, apply to your approval?"

"Who else needs to sign off before deployment?"

"Is there a phased rollout you'd prefer?"

Validation Report Template

# Validation Report: [Feature Name]

**Status**: Validated / Conditional / Failed
**Date**: [YYYY-MM-DD]
**Stakeholders**: [Names]
**Research Doc**: [Link to Phase 1]
**Design Doc**: [Link to Phase 2]
**Verification Report**: [Link to Phase 4]

## Executive Summary

[2-3 sentences on validation outcome]

## Specialist Skill Results

### Performance (`rust-performance` skill)
- **Benchmarks run**: [list]
- **Build profile**: release-lto
- **Targets met**: [Y/N with details]
- **Evidence**: [Criterion report, hyperfine results]

### Security (`security-audit` skill)
- **Scope**: [endpoints, auth flows audited]
- **OWASP findings**: [summary]
- **Critical issues**: [count]
- **Evidence**: [audit report link]

### Visual Regression (`visual-testing` skill) - if applicable
- **Surfaces covered**: [pages/components]
- **Baseline status**: [established/updated]
- **Regressions found**: [count]
- **Evidence**: [screenshot diffs]

### Acceptance Testing (`acceptance-testing` skill)
- **UAT Plan**: [link]
- **Scenarios executed**: [X/Y]
- **Pass rate**: [%]
- **Evidence**: [test results, recordings]

### Requirements Traceability (`requirements-traceability` skill)
- **Matrix location**: [path/link]
- **Requirements traced**: [X/Y]
- **Gaps**: [blockers/follow-ups]

### Quality Gate (`quality-gate` skill)
- **Decision**: Pass / Pass with Follow-ups / Fail
- **Report**: [link to Quality Gate Report]

## System Test Results

### End-to-End Scenarios

| ID | Workflow | Steps | Result | Status |
|----|----------|-------|--------|--------|
| E2E-001 | User Registration | 4 steps | All passed | PASS |
| E2E-002 | Data Export | 3 steps | All passed | PASS |

### Non-Functional Requirements (from specialist skills)

| Category | Target | Actual | Skill Used | Status |
|----------|--------|--------|------------|--------|
| Latency (p95) | < 100ms | 45ms | `rust-performance` | PASS |
| Memory | < 512MB | 380MB | `rust-performance` | PASS |
| Security Scan | No critical | 0 findings | `security-audit` | PASS |
| Accessibility | WCAG 2.1 AA | Compliant | Manual + axe | PASS |
| Visual Regression | No unintended | 0 diffs | `visual-testing` | PASS |

### NFR Details

[Detailed tables from specialist skill outputs]

## Acceptance Results (from `acceptance-testing` skill)

### Requirements Traceability (from `requirements-traceability` skill)

| Requirement ID | Description | Evidence | Stakeholder | Status |
|----------------|-------------|----------|-------------|--------|
| REQ-001 | User can register | E2E-001 passed | [Name] | Accepted |
| REQ-002 | Data export works | E2E-002 passed | [Name] | Accepted |

### Acceptance Interview Summary

**Date**: [YYYY-MM-DD]
**Participants**: [Names]

#### Problem Validation
[Summary of discussion - does it solve the problem?]

#### Success Criteria
[Summary - have criteria been met?]

#### Completeness
[Summary - anything missing?]

#### Risk Assessment
[Summary - deployment risks identified]

#### Conditions
[Any conditions attached to approval]

### Outstanding Concerns

| Concern | Raised By | Resolution | Status |
|---------|-----------|------------|--------|
| [Concern 1] | [Name] | [How resolved] | Resolved |

## Defect Register

| ID | Description | Origin Phase | Severity | Resolution | Status |
|----|-------------|--------------|----------|------------|--------|
| V001 | Slow under load | Phase 2 | High | Redesigned query | Closed |
| V002 | Missing audit log | Phase 1 | Medium | Added requirement | Closed |

## Sign-off

| Stakeholder | Role | Decision | Conditions | Date |
|-------------|------|----------|------------|------|
| [Name] | Product Owner | Approved | None | [Date] |
| [Name] | Security Lead | Approved | Quarterly re-scan | [Date] |
| [Name] | Ops Lead | Approved | Monitoring dashboard | [Date] |

## Gate Checklist

### Specialist Skill Outputs
- [ ] `rust-performance`: All benchmarks pass targets from research
- [ ] `security-audit`: No critical/high findings, or remediated
- [ ] `visual-testing`: No unintended regressions (if UI in scope)
- [ ] `acceptance-testing`: All UAT scenarios pass
- [ ] `requirements-traceability`: Matrix complete, no blocker gaps
- [ ] `quality-gate`: Final report shows Pass or Pass with Follow-ups

### Validation Gates
- [ ] All end-to-end workflows tested
- [ ] NFRs from research validated
- [ ] All requirements traced to acceptance evidence
- [ ] Stakeholder interviews completed
- [ ] All critical defects resolved (looped back and re-verified)
- [ ] Formal sign-off received from all required stakeholders
- [ ] Deployment conditions documented
- [ ] Ready for production

## Appendix

### Interview Transcript
[Detailed Q&A from acceptance sessions]

### Test Evidence
[Links to test reports, screenshots, recordings]

### Compliance Artifacts
[Any required compliance documentation]

Gate Criteria

Before production deployment:

Specialist Skill Requirements

  • rust-performance: Benchmarks pass all NFR targets from research
  • security-audit: No critical/high findings (or remediated and re-audited)
  • visual-testing: No unintended visual regressions (if UI in scope)
  • acceptance-testing: All UAT scenarios executed and passing
  • requirements-traceability: Complete matrix with all requirements traced
  • quality-gate: Final go/no-go report produced with Pass status

Core Validation Requirements

  • All user workflows tested end-to-end
  • NFRs from research validated (performance, security, accessibility)
  • All requirements traced to acceptance evidence
  • Stakeholder interviews completed using structured framework
  • All critical and high defects resolved through loop-back
  • Formal sign-off received from all required stakeholders
  • Deployment conditions documented and achievable
  • Ready for production deployment

Constraints

  • No skipping system tests: All NFRs must be verified
  • Structured interviews: Use AskUserQuestionTool framework
  • Trace to research: Every acceptance criterion links to Phase 1
  • Loop back properly: Defects go through left-side phases
  • Formal sign-off: No deployment without documented approval

Success Metrics

  • All requirements from research have acceptance evidence
  • All NFRs meet targets specified in research
  • Stakeholders formally approve for production
  • No critical or high defects open
  • Deployment conditions documented and achievable
  • Complete audit trail from requirement to acceptance

Deployment Readiness

After Phase 5 approval, the feature is ready for production deployment with:

  • Complete V-model traceability
  • Formal stakeholder sign-off
  • All defects resolved through proper phases
  • NFRs validated
  • Deployment conditions documented

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

36.96%
按下载量换算61

Claude

28.69%
按下载量换算48

Cursor

18.81%
按下载量换算31

Gemini CLI

9.76%
按下载量换算16

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills