changed skills and agent to ask skill to call agents

README.md

@@ -1,3 +1,158 @@
-# claude
+# claude-config
 
-Config de claude
+Global Claude Code configuration — agents, skills, and project templates.
+
+---
+
+## Overview
+
+This repo contains the global Claude Code setup used across all projects.
+
+```
+claude-config/
+├── CLAUDE.md              # Global coding preferences (style, rules, workflow)
+├── agents/                # Specialized agent definitions (called by skills or orchestrators)
+├── skills/                # Slash commands (/analyze, /debug, /ship-feature, ...)
+└── templates/
+    └── project-CLAUDE.md  # Template for per-project .claude/CLAUDE.md
+```
+
+**Architecture principle:**
+- `skills/` = entry points you invoke manually via `/skill-name`
+- `agents/` = execution units called by skills or by orchestrator agents
+- A skill delegates to one or more agents — it never contains logic itself
+
+---
+
+## Installation
+
+Clone the repo and symlink it into `~/.claude/`:
+
+```bash
+git clone git@github.com:youruser/claude-config.git ~/claude-config
+
+mkdir -p ~/.claude
+
+ln -sf ~/claude-config/agents    ~/.claude/agents
+ln -sf ~/claude-config/skills    ~/.claude/skills
+ln -sf ~/claude-config/CLAUDE.md ~/.claude/CLAUDE.md
+```
+
+Symlinks mean any update to this repo is immediately active — no manual sync needed.
+
+Verify the skills are loaded:
+
+```bash
+claude
+/skills
+```
+
+You should see all custom skills listed (`analyze`, `debug`, `ship-feature`, etc.).
+
+---
+
+## Available slash commands
+
+| Command | Description |
+|---|---|
+| `/analyze` | Deep analysis of code or a codebase before any modification |
+| `/architect` | Design a robust and scalable system architecture |
+| `/debug` | Find root cause and fix an issue precisely |
+| `/implement` | Implement a feature following project conventions |
+| `/refactor` | Improve code quality without changing behavior |
+| `/review` | Strict code review with severity-graded issues |
+| `/init-project` | Initialize a complete project from scratch (orchestrator) |
+| `/ship-feature` | Deliver a feature end-to-end via multi-agent pipeline (orchestrator) |
+
+Orchestrators (`/init-project`, `/ship-feature`) coordinate multiple agents sequentially with validation gates.
+Standalone skills (`/analyze`, `/debug`, etc.) invoke a single specialized agent.
+
+---
+
+## Agent pipeline (ship-feature)
+
+```
+/ship-feature <request>
+    └── ship-feature (orchestrator)
+            ├── analyzer    → understand the problem
+            ├── designer    → design the solution
+            ├── [validation gate — waits for user approval]
+            ├── implementer → write the code
+            ├── reviewer    → review loop (max 3 iterations)
+            └── tester      → define test strategy
+```
+
+---
+
+## Per-project setup
+
+Each project gets its own `.claude/CLAUDE.md` for local context and overrides.
+
+```bash
+# In your project root
+mkdir -p .claude
+cp ~/claude-config/templates/project-CLAUDE.md .claude/CLAUDE.md
+```
+
+Then fill in the relevant sections: build commands, test commands, conventions,
+architecture, and any exceptions to global rules.
+
+**Override rules:**
+- Local `.claude/` takes precedence over global `~/.claude/` for identical filenames
+- Files not defined locally fall back to global automatically
+- Use local overrides only for project-specific deviations — keep global rules generic
+
+---
+
+## Updating
+
+```bash
+cd ~/claude-config
+git pull
+```
+
+Changes are immediately active via symlinks. No restart needed for agents and skills
+(Claude Code reloads them at the start of each session).
+
+---
+
+## Adding a new skill or agent
+
+**New standalone skill** (single agent):
+
+1. Create `agents/myagent.md` — define role, tasks, rules, output format
+2. Create `skills/myskill.md`:
+
+```markdown
+---
+name: myskill
+description: One-line description of what this skill does
+argument-hint: <what to pass as argument>
+---
+
+Load and follow strictly:
+- .claude/agents/myagent.md
+
+Execute the MYAGENT agent on the following request:
+
+$ARGUMENTS
+```
+
+**New orchestrator skill** (multiple agents):
+
+1. Create `agents/myorchestrator.md` — define the workflow and agent call sequence
+2. Create `skills/myorchestrator.md` referencing all involved agents
+
+---
+
+## Extending per project
+
+If a project needs a modified version of an agent, place it in `.claude/agents/`:
+
+```bash
+# Override the implementer for a specific project
+cp ~/claude-config/agents/implementer.md .claude/agents/implementer.md
+# Edit to add project-specific constraints
+```
+
+The local version takes precedence. All other agents continue to load from global.

skills/analyze.md

@@ -1,23 +1,12 @@
-ROLE
-You are a senior software engineer specialized in code analysis.
+---
+name: analyze
+description: Analyze code or a codebase deeply before any modification
+argument-hint: <code, file, or area to analyze>
+---
 
-OBJECTIF
-Understand the provided code or project deeply before any modification.
+Load and follow strictly:
+- .claude/agents/analyzer.md
 
-CONTRAINTES
-- Do not modify anything
-- Do not assume missing behavior
-- Stay factual
+Execute the ANALYZER agent on the following target:
 
-PROCESS
-1. Identify the purpose of the code
-2. Identify main components (functions, modules, flows)
-3. Explain data flow
-4. Detect potential issues or unclear parts
-5. Highlight complexity or risks
-
-OUTPUT
-- Summary of what the code does
-- Key components
-- Data flow explanation
-- Risks / unclear areas
+$ARGUMENTS

skills/architect.md

@@ -0,0 +1,12 @@
+---
+name: architect
+description: Design a robust and scalable system architecture
+argument-hint: <system or feature to architect>
+---
+
+Load and follow strictly:
+- .claude/agents/architect.md
+
+Execute the ARCHITECT agent on the following request:
+
+$ARGUMENTS

skills/debug.md

@@ -1,24 +1,12 @@
-ROLE
-You are a senior engineer specialized in debugging.
+---
+name: debug
+description: Find root cause and fix an issue precisely
+argument-hint: <issue description or failing code>
+---
 
-OBJECTIF
-Find root cause and fix the issue.
+Load and follow strictly:
+- .claude/agents/debugger.md
 
-CONTRAINTES
-- Do not patch blindly
-- Identify root cause first
-- Preserve existing behavior outside the fix
+Execute the DEBUGGER agent on the following issue:
 
-PROCESS
-1. Understand expected behavior
-2. Reconstruct failing scenario
-3. Trace execution path
-4. Identify root cause
-5. Implement minimal fix
-6. Check for side effects
-
-OUTPUT
-- Root cause
-- Fix explanation
-- Code changes
-- Edge cases to watch
+$ARGUMENTS

skills/implement.md

@@ -1,25 +1,12 @@
-ROLE
-You are a senior engineer implementing a feature.
+---
+name: implement
+description: Implement a feature cleanly following project conventions
+argument-hint: <feature or change to implement>
+---
 
-OBJECTIF
-Implement the requested feature cleanly and safely.
+Load and follow strictly:
+- .claude/agents/implementer.md
 
-CONTRAINTES
-- Follow project conventions
-- Respect existing architecture
-- Preserve existing behavior
-- Follow strict code quality rules (see CLAUDE.md global)
+Execute the IMPLEMENTER agent on the following request:
 
-PROCESS
-1. Analyze existing code
-2. Identify integration points
-3. Propose minimal design
-4. Implement step by step
-5. Ensure no regression
-6. Keep code readable and modular
-
-OUTPUT
-- Explanation of approach
-- Code changes
-- Impacted parts
-- Potential risks
+$ARGUMENTS

skills/init-project.md

@@ -0,0 +1,17 @@
+---
+name: init-project
+description: Initialize a complete project from scratch with structure, stack, and base files
+argument-hint: <project idea or description>
+---
+
+Load and follow strictly:
+- .claude/agents/init-project.md
+- .claude/agents/analyzer.md
+- .claude/agents/designer.md
+- .claude/agents/implementer.md
+- .claude/agents/reviewer.md
+- .claude/agents/tester.md
+
+Execute the orchestrator defined in .claude/agents/init-project.md with the following request:
+
+$ARGUMENTS

skills/refactor.md

@@ -1,24 +1,12 @@
-ROLE
-You are a senior engineer performing a safe refactor.
+---
+name: refactor
+description: Improve code quality without changing behavior
+argument-hint: <file, function, or module to refactor>
+---
 
-OBJECTIF
-Improve code quality without changing behavior.
+Load and follow strictly:
+- .claude/agents/refactorer.md
 
-CONTRAINTES
-- Zero functional change
-- Follow global coding standards
-- Remove legacy residue
-- Respect project conventions
+Execute the REFACTORER agent on the following target:
 
-PROCESS
-1. Identify problems (complexity, duplication, readability)
-2. Propose refactor plan
-3. Apply small safe transformations
-4. Ensure behavior is unchanged
-5. Remove outdated or unused code
-
-OUTPUT
-- What was improved
-- Before/after reasoning
-- Any deviations from standards (with justification)
-- Confirmation that no old logic remains
+$ARGUMENTS

skills/review.md

@@ -1,25 +1,12 @@
-ROLE
-You are a strict but pragmatic code reviewer.
+---
+name: review
+description: Strict code review with severity-graded issues
+argument-hint: <file, function, or code to review>
+---
 
-OBJECTIF
-Evaluate code quality and compliance with standards.
+Load and follow strictly:
+- .claude/agents/reviewer.md
 
-CONTRAINTES
-- Apply global coding standards
-- Be precise, not verbose
-- Distinguish critical vs minor issues
+Execute the REVIEWER agent on the following code:
 
-PROCESS
-1. Check structure and readability
-2. Check function size and responsibilities
-3. Check parameters and variables
-4. Check naming clarity
-5. Detect code smells
-6. Identify deviations from standards
-7. Evaluate justification of deviations
-
-OUTPUT
-- Critical issues
-- Improvements
-- Deviations from standard
-- Justified vs unjustified deviations
+$ARGUMENTS

skills/ship-feature.md

@@ -0,0 +1,17 @@
+---
+name: ship-feature
+description: Ship a feature end-to-end via multi-agent orchestrator
+argument-hint: <description de la feature à implémenter>
+---
+
+Load and follow strictly these agent files:
+- .claude/agents/ship-feature.md
+- .claude/agents/analyzer.md
+- .claude/agents/designer.md
+- .claude/agents/implementer.md
+- .claude/agents/reviewer.md
+- .claude/agents/tester.md
+
+Execute the orchestrator defined in .claude/agents/ship-feature.md with the following request:
+
+$ARGUMENTS