Claude Code Plugins

Community-maintained marketplace

Feedback

This skill should be used when the user asks to "validate requirements", "review requirements quality", "check requirements completeness", "verify traceability", "check INVEST compliance", "validate user stories", "requirements health check", "quality gate check", or when running /re:review validation.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name validation
description This skill should be used when the user asks to "validate requirements", "review requirements quality", "check requirements completeness", "verify traceability", "check INVEST compliance", "validate user stories", "requirements health check", "quality gate check", or when running /re:review validation.

Validation

Quick Actions & Routing

User Intent Action Resource
Run full validation Execute /re:review command Command handles orchestration
Check completeness Verify all required elements references/completeness-checks.md
Check consistency Verify traceability and alignment references/consistency-checks.md
Validate stories (INVEST) Apply INVEST criteria references/invest-criteria.md
Understand thresholds Review pass/warning/fail levels references/quality-thresholds.md
Generate report Use standard format references/report-template.md
Fix common issues Apply fix patterns references/fix-patterns.md
View example report Load sample validation output examples/example-validation-report.md

Command Integration

The /re:review command orchestrates requirements validation in GitHub Projects. This skill provides the methodology for what to validate and how—including the four-dimensional validation framework, quality thresholds, and fix patterns. Load this skill for deeper understanding of validation concepts or when you need guidance beyond what the command provides.

Overview

Requirements validation ensures that requirements at all levels (vision, epics, stories, tasks) are complete, consistent, high-quality, and traceable. Validation identifies issues early, reducing rework and improving implementation success. This skill provides the framework for systematic validation.

Purpose

Validation serves as the quality gate for requirements:

  • Catches issues early: Before implementation begins
  • Ensures traceability: Vision → Epics → Stories → Tasks chain is complete
  • Verifies quality: INVEST criteria for stories, clear acceptance criteria
  • Identifies gaps: Missing elements, broken links, incomplete coverage

Four-Dimensional Validation Framework

Perform validation across four dimensions:

Dimension Question Focus
Completeness Are all required elements present? Structure and content
Consistency Do requirements align and link properly? Relationships and terminology
Quality Do requirements meet best practice standards? INVEST, acceptance criteria
Traceability Can we trace from vision to tasks? Parent/child hierarchy

Validation Process Overview

Step 1: Gather Requirements

Retrieve all items from GitHub Project:

gh project item-list [project-number] --format json

Categorize by Type: Vision, Epic, Story, Task. For each item, read full content:

gh issue view [issue-number] --repo [repo] --json body,title,labels

Step 2: Apply Completeness Checks

Verify required elements exist at each level. See references/completeness-checks.md for detailed checklists.

Quick summary by level:

Level Key Elements
Vision Problem, users, solution, metrics, scope
Epic Overview, value, scope, success criteria, parent link
Story Story format, acceptance criteria (3-5), parent link, size (1-5 days)
Task Action title, description, acceptance criteria (3-5), parent link, size (2-8 hrs)

Step 3: Apply Consistency Checks

Verify alignment and traceability. See references/consistency-checks.md for detailed checks.

Key checks:

  • Every epic links to vision
  • Every story links to an epic
  • Every task links to a story
  • No orphaned issues
  • Consistent terminology and labels
  • Child priorities don't exceed parent priorities

Step 4: Apply Quality Checks

Validate requirements meet quality standards. See references/invest-criteria.md for INVEST details.

INVEST criteria for stories:

Letter Criterion Quick Check
I Independent Can complete without other stories?
N Negotiable Implementation details open?
V Valuable Clear user/business value?
E Estimable Team can estimate effort?
S Small Fits in 1-5 days?
T Testable Specific acceptance criteria?

Acceptance criteria quality:

  • Specific and unambiguous
  • Testable and verifiable
  • Observable outcomes
  • Minimum 3-5 per story/task

Step 5: Generate Validation Report

Format findings using the standard report template. See references/report-template.md.

Report sections:

  1. Executive Summary (dimensions + overall status)
  2. Requirements Inventory (counts by level)
  3. Critical Issues (must fix)
  4. Warnings (should address)
  5. INVEST Compliance (for stories)
  6. Coverage Analysis
  7. Recommendations (prioritized)
  8. Next Steps

Step 6: Offer to Fix Issues

After presenting report, offer assistance:

  • Auto-fix: Automatically fix where possible
  • Guided fix: Walk through each issue
  • Skip: Review only, no changes

See references/fix-patterns.md for common fix approaches.

Quality Thresholds

Use these thresholds for pass/warning/fail assessment. See references/quality-thresholds.md for details.

Metric Pass Warning Fail
Completeness >90% 70-90% <70%
Consistency 100% 95-99% <95%
INVEST Compliance >80% 60-80% <60%
Traceability 100% 95-99% <95%
Acceptance Criteria >=3 per item 2 per item <2 per item

Issue Severity Classification

Critical Issues (Must Fix)

Block progress and must be resolved:

  • Missing vision (no vision issue exists)
  • Broken traceability (orphaned items without parents)
  • Missing acceptance criteria (cannot verify completion)
  • Incomplete scope definitions

Warnings (Should Address)

Quality issues that should be fixed but don't block:

  • Oversized stories (>5 days)
  • INVEST violations
  • Priority imbalances (>60% Must Have)
  • Vague task descriptions

Best Practices

Be Thorough but Pragmatic

  • Focus on actionable findings
  • Distinguish critical from nice-to-have
  • Don't be pedantic about minor style issues

Provide Actionable Guidance

  • Every issue should have a clear fix
  • Reference specific issue numbers
  • Group recommendations by priority

Validate Iteratively

  • Re-run validation after fixes
  • Use as quality gate before sprint planning
  • Recommend periodic reviews (weekly/monthly)

Reference Files

Load references as needed:

Reference When to Load Path
completeness-checks.md Detailed per-level checklists references/completeness-checks.md
consistency-checks.md Traceability and alignment checks references/consistency-checks.md
invest-criteria.md INVEST criteria deep-dive references/invest-criteria.md
quality-thresholds.md Pass/warning/fail thresholds references/quality-thresholds.md
report-template.md Validation report format references/report-template.md
fix-patterns.md Common fixes for common issues references/fix-patterns.md

Examples

Working examples that can be copied and adapted:

Example Use Case Path
example-validation-report.md Viewing a complete validation report with realistic findings examples/example-validation-report.md

Related Skills

Load these skills when validation reveals needs beyond this skill's scope:

Validation Finding Load Skill Routing Trigger
Vision is missing or incomplete vision-discovery Need to create or improve vision
Epics are missing or poorly defined epic-identification Need to identify or refine epics
Stories fail INVEST criteria user-story-creation Need to rewrite or split stories
Tasks are missing or oversized task-breakdown Need to create or break down tasks
Priorities are imbalanced prioritization Need to apply MoSCoW framework