Menu
Add specialized craft skills to agents. Default craft principles always active. Everything MUST respect the craft philosophy.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
Craft something. Claude orchestrates, agents execute.
Start a Clean Claude agent with optional reactive links. Examples: /agent frontend-engineer, /agent frontend-engineer --link qa-engineer, /agent architect --link frontend-engineer,qa-engineer
Auto-repair with smart routing: test failures → Dev, type errors → Architect, spec gaps → PO. Routes each problem to the right expert.
Re-run stack detection and skill generation. Use when stack evolved or on first run.
Set up the Clean Claude Reactive System in the current project. Configures hooks, shared state, and scripts for the multi-agent feedback loop
Bootstrap a new frontend project with craft principles: React + Vite + TypeScript + Vitest + clean architecture
| name | add-skill |
| description | Add specialized craft skills to agents. Default craft principles always active. Everything MUST respect the craft philosophy. |
| user-invocable | false |
| context | conversation |
| allowed-tools | Read, Write, Bash, Task, AskUserQuestion |
Craft principles are ALWAYS active. Add specialized expertise on top.
┌─────────────────────────────────────────────────────────────────┐
│ │
│ 🏛️ CRAFT FOUNDATION (Always Active) │
│ ────────────────────────────────────────── │
│ │
│ • SOLID Principles — Single responsibility, Open/closed, etc. │
│ • Clean Architecture — Dependencies point inward │
│ • Hexagonal (Ports & Adapters) — Domain at center │
│ • Explicit Error Handling — Result<T, E>, no throw │
│ • Type Safety — Strict mode, no any │
│ • Immutability — Const by default │
│ • Tested — Code always has tests (BDD style) │
│ │
└─────────────────────────────────────────────────────────────────┘
/add-skill <agent> <skill> [--scope <path>]
/add-skill --list # Show all active skills
/add-skill architect DDD # Add DDD expertise to architect
/add-skill architect CQRS # Add CQRS patterns
/add-skill architect CQRS --scope src/orders # Scoped to a path
/add-skill backend-engineer Event-Sourcing
/add-skill frontend-engineer React-Patterns
/add-skill qa-engineer Property-Testing
/add-skill --list # See what's active
Every skill MUST respect the craft philosophy.
When you add a skill, Clean Claude validates it:
┌─────────────────────────────────────────────────────────────────┐
│ │
│ /add-skill architect DDD │
│ │
│ ✅ VALIDATION │
│ │
│ 1. Is DDD craft-aligned? │
│ ✓ Domain at center — aligns with Hexagonal │
│ ✓ Bounded contexts — aligns with Clean Architecture │
│ ✓ Aggregates — aligns with Single Responsibility │
│ ✓ Value Objects — aligns with Immutability │
│ │
│ 2. Does it conflict with existing skills? │
│ ✓ No conflicts │
│ │
│ ✅ APPROVED: DDD is craft-aligned │
│ │
└─────────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────────────────────────┐
│ │
│ /add-skill architect "Throw exceptions everywhere" │
│ │
│ ❌ REJECTED │
│ │
│ This skill violates craft principles: │
│ ✗ Conflicts with: Explicit Error Handling │
│ ✗ Reason: throw is implicit, Result<T, E> is explicit │
│ │
│ Suggestion: Use "Explicit-Errors" skill instead │
│ │
└─────────────────────────────────────────────────────────────────┘
All these skills have been validated against the craft philosophy:
| Skill | Description | Craft Alignment |
|---|---|---|
DDD | Domain-Driven Design | ✅ Domain-first, bounded contexts |
CQRS | Command Query Responsibility Segregation | ✅ Single responsibility, explicit |
Event-Sourcing | Event store, projections, replay | ✅ Immutability, explicit state |
Hexagonal-Advanced | Deep ports & adapters patterns | ✅ Extension of core Hexagonal |
Modular-Monolith | Module boundaries, internal APIs | ✅ Bounded contexts, clean deps |
| Skill | Description | Craft Alignment |
|---|---|---|
React-Patterns | Compound components, custom hooks | ✅ Composition, single responsibility |
State-Machines | XState, statecharts | ✅ Explicit state, predictable |
A11y-Expert | Accessibility deep-dive | ✅ Domain (user needs) first |
Performance | Core Web Vitals optimization | ✅ Measurable, testable |
| Skill | Description | Craft Alignment |
|---|---|---|
API-Design | REST best practices, HATEOAS | ✅ Explicit contracts, testable |
GraphQL | Schema design, resolvers | ✅ Type-safe, explicit |
Message-Queues | Event-driven patterns | ✅ Decoupling, explicit |
Caching | Cache strategies, invalidation | ✅ Explicit, predictable |
| Skill | Description | Craft Alignment |
|---|---|---|
Testing-Advanced | Mocking strategies, test doubles | ✅ Extension of core testing |
BDD | Gherkin, behavior specs | ✅ Domain language, explicit |
Contract-Testing | Pact, consumer-driven | ✅ Explicit contracts |
Property-Testing | QuickCheck-style, fuzzing | ✅ Exhaustive, type-driven |
┌─────────────────────────────────────────────────────────────────┐
│ /add-skill architect DDD │
│ │
│ 1. CRAFT VALIDATION │
│ ───────────────── │
│ → Is DDD craft-aligned? ✅ │
│ → Any conflicts? None ✅ │
│ │
│ 2. SKILL GENERATION │
│ ───────────────────── │
│ → Use architect agent to generate DDD skill document │
│ → Include: principles, patterns, examples, checklist │
│ → Store in .clean-claude/skills/ddd.md │
│ │
│ 3. CODEBASE ANALYSIS │
│ ──────────────────── │
│ → Architect analyzes code through DDD lens │
│ → Identifies: bounded contexts, aggregates, violations │
│ → Generates refactoring plan │
│ │
│ 4. USER APPROVAL │
│ ───────────────── │
│ → Show analysis and plan │
│ → User approves/modifies/rejects │
│ │
│ 5. REACTIVE LOOP (if approved) │
│ ────────────────────────── │
│ → Architect designs changes │
│ → Engineers implement │
│ → QA verifies │
│ → Loop until green │
│ │
└─────────────────────────────────────────────────────────────────┘
> /add-skill --list
┌─────────────────────────────────────────────────────────────────┐
│ ACTIVE SKILLS │
│ │
│ 🏛️ CRAFT FOUNDATION (Always Active) │
│ • SOLID │
│ • Clean Architecture │
│ • Hexagonal │
│ • Explicit Errors │
│ • Type Safety │
│ • Immutability │
│ • Tested │
│ │
│ ➕ ADDED SKILLS │
│ • architect: DDD (added 2024-01-15) │
│ • architect: CQRS (added 2024-01-16) │
│ • qa-engineer: Property-Testing (added 2024-01-16) │
│ │
└─────────────────────────────────────────────────────────────────┘
.clean-claude/
├── context.json # Includes active skills
├── skills/
│ ├── ddd.md # Generated skill document
│ ├── cqrs.md # Generated skill document
│ └── ...
└── refactoring-plans/
└── ddd-2024-01-15.md # Refactoring plan if applied
{
"stack": { ... },
"craftFoundation": [
"SOLID",
"Clean-Architecture",
"Hexagonal",
"Explicit-Errors",
"Type-Safety",
"Immutability",
"Tested"
],
"addedSkills": [
{
"name": "DDD",
"agent": "architect",
"addedAt": "2024-01-15T10:30:00Z",
"scope": null
},
{
"name": "CQRS",
"agent": "architect",
"addedAt": "2024-01-16T09:00:00Z",
"scope": "src/features/orders"
}
]
}
When agents are spawned, they receive:
┌─────────────────────────────────────────────────────────────────┐
│ │
│ AGENT CONTEXT = Foundation + Stack + Patterns + Skills │
│ │
│ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ ┌─────────┐ │
│ │ SOLID │ │ React │ │ Your │ │ DDD │ │
│ │ Clean │ │ Query │ │ naming │ │ CQRS │ │
│ │ Hexa │ │ patterns │ │ patterns │ │ etc. │ │
│ │ Tested │ │ │ │ │ │ │ │
│ └─────────────┘ └─────────────┘ └─────────────┘ └─────────┘ │
│ Foundation Stack Defaults Learned Added │
│ (Always) (Generated) (If clean) (Manual) │
│ │
└─────────────────────────────────────────────────────────────────┘
type Command =
| { type: 'add'; agent: string; skill: string; scope?: string }
| { type: 'list' };
// /add-skill architect DDD → { type: 'add', agent: 'architect', skill: 'DDD' }
// /add-skill --list → { type: 'list' }
Task(
subagent_type: "architect",
prompt: """
Validate if this skill is CRAFT-ALIGNED:
Skill: DDD (Domain-Driven Design)
Target Agent: architect
Check against craft foundation:
1. Does it align with SOLID?
2. Does it align with Clean Architecture?
3. Does it align with Hexagonal?
4. Does it align with Explicit Errors?
5. Does it align with Type Safety?
6. Does it align with Immutability?
7. Does it include tests?
If ANY conflict → REJECT with explanation
If aligned → APPROVE with alignment notes
"""
)
Task(
subagent_type: "architect",
prompt: """
Generate an expert-level skill document for: DDD
Structure:
## Core Principles
## Tactical Patterns (with TypeScript examples)
## Strategic Patterns
## Integration with Craft Foundation
## Checklist for Code Review
## Anti-patterns to Avoid
IMPORTANT: All examples must use Result<T, E>, strict types, etc.
"""
)
Same reactive flow as before — analyze codebase, propose plan, loop.
> /add-skill --list
🏛️ CRAFT FOUNDATION (Always Active)
• SOLID • Clean Architecture • Hexagonal
• Explicit Errors • Type Safety • Immutability • Tested
➕ ADDED SKILLS
(none)
> /add-skill architect DDD
✅ Validating DDD against craft principles...
✓ Aligns with SOLID (aggregates = single responsibility)
✓ Aligns with Clean Architecture (domain at center)
✓ Aligns with Hexagonal (bounded contexts as ports)
📝 Generating DDD skill document...
✓ Stored in .clean-claude/skills/ddd.md
🔍 Analyzing codebase with DDD lens...
📋 Analysis Complete
## Bounded Contexts Identified
- Users (current)
- Orders (current)
- Payments (referenced)
## Violations Found
| File | Issue | Priority |
|------|-------|----------|
| UserService.ts | Anemic domain model | 🔴 High |
| Order.ts | No aggregate boundaries | 🔴 High |
## Refactoring Plan
1. **Phase 1**: Create User aggregate
2. **Phase 2**: Create Order aggregate with items
─────────────────────────────
[Approve & Implement] [Modify Scope] [Just add skill, don't refactor] [Cancel]
> Just add skill, don't refactor
✅ DDD added to architect
> /add-skill --list
🏛️ CRAFT FOUNDATION (Always Active)
• SOLID • Clean Architecture • Hexagonal
• Explicit Errors • Type Safety • Immutability • Tested
➕ ADDED SKILLS
• architect: DDD