agentic-executor

# Agentic Executor Skill

**Reusable capabilities for autonomous multi-file operations.**

## Purpose

This skill provides reusable agentic capabilities for:
- Multi-file refactoring
- Parallel execution coordination
- Impact analysis
- Safe rollback mechanisms

## When to Use This Skill

Invoke this skill when you need:
1. Autonomous multi-file operations
2. Parallel processing with conflict detection
3. Safe refactoring with rollback
4. Impact analysis and preview

## Core Capabilities

### 1. Impact Analysis

**Analyze affected files before making changes:**

```bash
# Function: analyze_impact
# Input: Task description, transformation type
# Output: List of affected files, dependency graph

analyze_impact() {
  local task="$1"
  local type="$2"  # RENAME, EXTRACT, RESTRUCTURE

  echo "🔍 Analyzing impact for: $task"

  # Discover affected files
  case "$type" in
    RENAME)
      grep -r "class $old_name" --include="*.ts" --include="*.js" -l
      ;;
    EXTRACT)
      grep -r "$pattern" --include="*.ts" -l
      ;;
    RESTRUCTURE)
      find "$target_dir" -type f -name "*.ts"
      ;;
  esac
}
```

**Returns:**
- List of affected files
- Dependency relationships
- Risk assessment (LOW/MEDIUM/HIGH)

### 2. Dependency Resolution

**Build dependency graph and determine execution order:**

```python
# Function: resolve_dependencies
# Input: List of files, their imports
# Output: Execution levels (batches)

def resolve_dependencies(files):
    levels = {}
    visited = set()

    def get_level(file):
        if file in visited:
            return levels.get(file, 0)

        visited.add(file)
        max_dep_level = 0

        for dep in get_imports(file):
            dep_level = get_level(dep)
            max_dep_level = max(max_dep_level, dep_level + 1)

        levels[file] = max_dep_level
        return max_dep_level

    for file in files:
        get_level(file)

    return levels
```

**Example Output:**
```
Level 0: [User.ts]  # No dependencies
Level 1: [Use