nw-ab-critique-dimensi…
Review dimensions for validating agent quality - template compliance, safety, testing, and priority validation
Definition of Ready checklist criteria, antipattern detection patterns, UAT quality rules, and domain language enforcement for product owner review
$ npx -y skills add nWave-ai/nWave --skill nw-dor-validation --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/nw-dor-validationContext preview
The summary Claude sees to decide when to auto-load this skill.
Definition of Ready checklist criteria, antipattern detection patterns, UAT quality rules, and domain language enforcement for product owner review
name: nw-dor-validation description: Definition of Ready checklist criteria, antipattern detection patterns, UAT quality rules, and domain language enforcement for product owner review user-invocable: false disable-model-invocation: true
All items must PASS with evidence. Each FAIL blocks handoff.
Domain language (not technical), describes real user pain, testable. Pass: "Maria wastes 30 seconds typing credentials on every visit" Fail: "Users need authentication" | "Implement login feature"
Real name, specific role, clear context. Pass: "Maria Santos, returning customer (2+ orders), using trusted MacBook" Fail: "User" | "Customer" | "End user" | "Authenticated user"
Min 3 examples, real names (not user123), real values, different scenarios (happy/edge/error). Pass: "Example 1: Maria on MacBook, 5 days since login, goes to dashboard" Fail: "User logs in successfully" | "Test with valid credentials"
Given/When/Then format, 3-7 scenarios, real data, covers happy + edge. Pass: "Given Maria authenticated on 'MacBook-Home' 5 days ago..." Fail: "Test login works" | "Given a user When they login Then success"
Checkable (checkbox), traceable to UAT, outcome-focused (not implementation). Pass: "Sessions older than 30 days require re-authentication" Fail: "Use JWT tokens" | "System should work correctly"
Effort estimate provided, scenario count in range, single demonstrable outcome. Pass: 2 days, 5 UAT scenarios, demoed in single session Fail: >7 scenarios | >3 days | multiple distinct outcomes
Dependencies listed, risks identified, architectural considerations noted. Pass: "Requires JWT token storage, GDPR cookie consent integration" Fail: no technical notes section
Blocking deps identified, resolution status clear, escalation path. Pass: "Depends on US-041 (completed) and Auth service API (available)" Fail: "Needs some API - TBD"
---
Signal: starts with "Implement", "Add", "Create", "Build", "Develop" Detection: `^(Implement|Add|Create|Build|Develop)\s` | Fix: rewrite as user pain
Signal: user123, test@test.com, foo, bar, lorem, placeholder Detection: `user[0-9]+`, `test@`, `example@`, `foo`, `bar` | Fix: real names -- Maria Santos
Signal: AC describes implementation not outcome Detection: "Use JWT", "Implement using", "Database should", "API must return" Fix: outcome focus -- "Session persists for 30 days"
Signal: >7 scenarios | >3 days | multiple distinct outcomes | Fix: split by user outcome
Signal: no "Example" section | <3 examples | abstract examples | Fix: add 3+ with real data
Signal: "Tests to be added", "Will write tests later", "Tests TBD" | Fix: UAT first, RED first
Signal: "User", "Customer", "End user" as persona | Fix: "Maria Santos, returning customer (2+ orders)"
Signal: all success scenarios, no errors, no boundaries | Fix: add expired session, invalid device, etc.
---
**Format**: Given/When/Then with complete sentences. Fail: "Test login", "Given user When login Then success" **Real Data**: real names, values, scenarios. Fail: "Given user123", "When X happens" **Coverage**: min 1 happy path + 1 edge + 1 error. Range: 3-7 scenarios.
---
**Technical Jargon**: flag in user-facing sections: JWT, API, database, backend, frontend, microservice, REST, HTTP, JSON, SQL. Exception: Technical Notes section. Fix: "session token" -> "remember me"
**Generic Language**: flag "the system", "the application", "functionality", "feature". Fix: use specific names -- "the login page" -> "the welcome screen"
AI agents that guide you from idea to working code, with human judgment at every gate. nWave runs inside Claude Code. It breaks feature delivery into seven waves (discover, diverge, discuss, design, devops, distill, deliver).
Repo: nWave-ai/nWave
Review dimensions for validating agent quality - template compliance, safety, testing, and priority validation
Review dimensions for validating agent quality - template compliance, safety, testing, and priority validation
Review dimensions for acceptance test quality - happy path bias, GWT compliance, business language purity, coverage completeness, walking skeleton…
Detailed 5-phase workflow for creating agents - from requirements analysis through validation and iterative refinement
5-layer testing approach for agent validation including adversarial testing, security validation, and prompt injection resistance
Architectural style selection decision matrices, trade-off analysis, structural enforcement rules, and combination patterns. Load when choosing or evaluating…