BYAN CLAUDE.md
Des agent IA mais intelligent qui respect toute les bonne pratique de conception
> Sourced from [Yan-Acadenice/BYAN](https://github.com/Yan-Acadenice/BYAN) — [MIT](https://github.com/Yan-Acadenice/BYAN/blob/60924ba300e22498854662d748ec611ac4fa27cf/.claude/agents/bmad-claude.md).
# BMAD Platform — Instructions Claude Code
Ce fichier est charge automatiquement par Claude Code pour toutes les interactions dans ce projet.
## Plateforme
**BMAD (Business Modeling & Agent Development)** — Plateforme modulaire d'agents IA specialises orchestres via des workflows structures. Methodologie Merise Agile + TDD, 64 mantras fondamentaux.
## Modules
| Module | Chemin | Role |
|--------|--------|------|
| **Core** | `_bmad/core/` | Fondation : party-mode, brainstorming, taches de base |
| **BMM** | `_bmad/bmm/` | Cycle de dev complet : Analyse → Planning → Solution → Implementation |
| **BMB** | `_bmad/bmb/` | Meta-systeme : creation d'agents, modules, workflows |
| **TEA** | `_bmad/tea/` | Architecture de tests : ATDD, automation, CI/CD, NFR |
| **CIS** | `_bmad/cis/` | Innovation creative : design thinking, storytelling |
## Architecture des Agents
**Format** : Fichiers Markdown avec frontmatter YAML + definitions XML
**Emplacement** : `_bmad/{module}/agents/{agent-name}.md`
**Structure** : Frontmatter → Activation → Persona → Menu → Knowledge Base → Capabilities
**Manifeste** : `_bmad/_config/agent-manifest.csv`
### Soul System (TAO)
Chaque agent peut avoir une ame (soul.md) et une voix (tao.md) :
- **Soul** : Personnalite, lignes rouges, rituels, phrase fondatrice
- **Tao** : Registre vocal, signatures verbales, temperature emotionnelle
- **Soul-Memory** : Journal vivant des sessions passees
- **Protocole** : `_byan/core/activation/soul-activation.md`
## Capacites Natives (Tous les Agents)
Tous les agents BMAD sont **nativement** workflow-aware et delegation-capable.
### Invoquer un Workflow
```
@bmad-{module}-{workflow}
Exemple : @bmad-bmm-create-prd, @bmad-tea-testarch-atdd
```
### Deleguer a un Agent
```
@bmad-agent-{name}
Exemple : @bmad-agent-bmm-dev, @bmad-agent-byan
```
### Variables de Contexte
- `{project-root}` : Racine du repository
- `{output_folder}` : `_bmad-output/`
- `{planning_artifacts}` : `_bmad-output/planning-artifacts/`
- `{implementation_artifacts}` : `_bmad-output/implementation-artifacts/`
- `{user_name}`, `{communication_language}` : Depuis config.yaml
### Orchestration Multi-Agent
- Les agents peuvent **invoquer d'autres agents** en cours de workflow
- Les agents peuvent **lancer des workflows paralleles** pour des taches complexes
- **Party Mode** (`@bmad-party-mode`) pour discussions multi-agents
### Menu Handlers
Les agents executent des actions via des handlers de menu :
- `exec` : Executer un fichier/workflow
- `workflow` : Lancer un workflow multi-etapes
- `tmpl` : Generer depuis un template
- `data` : Charger des donnees contextuelles
- `action` : Action inline
- `validate-workflow` : Valider un workflow
## Workflows
**Format** : Markdown multi-etapes avec fichiers dans `steps/`
**Emplacement** : `_bmad/{module}/workflows/{workflow-name}/workflow.{md|yaml}`
**Manifeste** : `_bmad/_config/workflow-manifest.csv`
**Types** :
- **Tri-modal** : Create / Validate / Edit (PRD, Architecture, Agents)
- **Sequentiel** : Processus guide multi-phases (Interview, Sprint Planning)
- **Tache** : Utilitaire ponctuel (Editorial Review, Shard Doc)
## Phases de Developpement
### Phase 0 : Setup
`@bmad-bmm-document-project` → `@bmad-bmm-generate-project-context`
### Phase 1 : Analyse
`@bmad-brainstorming` → `@bmad-bmm-create-brief` → `@bmad-bmm-research`
### Phase 2 : Planning
`@bmad-bmm-create-prd` → `@bmad-bmm-create-ux-design` → `@bmad-bmm-validate-prd`
### Phase 3 : Solution
`@bmad-bmm-create-architecture` → `@bmad-bmm-create-epics-and-stories` → `@bmad-bmm-check-implementation-readiness`
### Phase 4 : Implementation
`@bmad-bmm-sprint-planning` → `@bmad-bmm-create-story` → `@bmad-bmm-dev-story` → `@bmad-bmm-code-review`
### Quick Flow (Brownfield)
`@bmad-bmm-quick-spec` → `@bmad-bmm-quick-dev`
## Conventions
### Git
- Format : `<type>: <description>` (feat, fix, docs, refactor, test, chore)
- **ZERO emoji** dans les commits (Mantra IA-23)
### Code
- Self-documenting, commentaires uniquement pour le POURQUOI
- Fonctions font une seule chose, max 3 parametres
- Pas d'effets de bord dans les fonctions pures
### Anti-Patterns
- Ne jamais accepter sans questionner (IA-16 Challenge Before Confirm)
- Ne jamais assumer que l'utilisateur a raison (IA-1 Zero Trust)
- Ne jamais ajouter de features "au cas ou" (YAGNI)
- Toujours evaluer les consequences avant d'agir (#39)
## Structure du Projet
```
{project-root}/
├── _bmad/ # Code de la plateforme
│ ├── _config/ # Manifestes (agents, workflows, tasks)
│ ├── _memory/ # Memoire persistante des agents
│ ├── core/ # Module fondation
│ │ └── activation/ # Protocoles d'activation (soul-activation.md)
│ ├── bmm/ # Module SDLC
│ ├── bmb/ # Module Builder
│ ├── tea/ # Module Test Architecture
│ └── cis/ # Module Innovation
├── _byan/ # Soul System BYAN
│ ├── agents/ # Agents principaux
│ ├── soul.md # Ame de BYAN
│ ├── tao.md # Voix de BYAN
│ ├── soul-memory.md # Journal vivant
│ └── config.yaml # Configuration
├── _bmad-output/ # Artefacts generes
├── .github/agents/ # Stubs Copilot CLI
└── .codex/prompts/ # Stubs Codex
```
## References
- Agent Manifest : `_bmad/_config/agent-manifest.csv`
- Workflow Manifest : `_bmad/_config/workflow-manifest.csv`
- Task Manifest : `_bmad/_config/task-manifest.csv`
- Soul Activation : `_byan/core/activation/soul-activation.md`
- README : `README.md`
Add to your project
Paste into your project's CLAUDE.md or ~/.claude/CLAUDE.md for global rules.
More for Rust
Node.js Express API
by @Claude Rules
Building scalable Node.js REST APIs with Express, middleware, and proper async patterns.
Rust Systems Programming
by @Claude Rules
Safe, fast Rust code leveraging the ownership system, traits, and zero-cost abstractions.
Coolify Docs CLAUDE.md
by @coollabsio
Documentation for Coolify
Awesome Claude Notes CLAUDE.md
by @loulanyue
🏄 帮助你快速搭建 Claude Code 与 AI Agent 生产力工作流的实战仓库 🏄 A practical toolkit to help you quickly build high-productivity workflows for Claude Code and AI agents
N8n Install CLAUDE.md
by @kossakovsky
🚀 Self-hosted AI automation platform. Deploy n8n, Ollama, Flowise, RAG, Supabase & 30+ tools with one command. Auto HTTPS. Free Zapier/Make alternative.
Cc Plugin Catalog CLAUDE.md
by @giginet
Static site generator for Claude Code Plugin Marketplace repositories
MCP servers for Rust
netdata/netdata#Netdata
🎖️ 🏠 ☁️ 📟 🍎 🪟 🐧 - Discovery, exploration, reporting and root cause analysis using all observability data, including metrics, logs, systems, containers, processes, and network connections
eyaltoledano/claude-task-master
📇 ☁️ 🏠 - AI-powered task management system for AI-driven development. Features PRD parsing, task expansion, multi-provider support (Claude, OpenAI, Gemini, Perplexity, xAI), and selective tool loadi
mediar-ai/screenpipe
🎖️ 🦀 🏠 🍎 Local-first system capturing screen/audio with timestamped indexing, SQL/embedding storage, semantic search, LLM-powered history analysis, and event-triggered actions - enables building c
Browse by Tag
Get the Claude Code Starter Pack
Top CLAUDE.md rules for Next.js, TypeScript, Python, Go, and React — free.
