- name
- hooks-automation
- description
- Configure Claude Code hooks for automated coordination and learning
- version
- 1.0.0
- invocable
- true
- author
- agentic-flow
- capabilities
- ["pre_edit_hooks","post_edit_hooks","session_management","metrics_tracking"]
# Hooks Automation Skill
Configure intelligent hooks for automated coordination, formatting, and learning.
## Available Hooks
### PreToolUse Hooks
Triggered before Claude Code tools execute.
```json
{
"PreToolUse": [
{
"matcher": "Edit|Write|MultiEdit",
"hooks": [{
"type": "command",
"command": "npx agentic-flow@alpha hooks pre-edit \"$TOOL_INPUT_file_path\""
}]
},
{
"matcher": "Bash",
"hooks": [{
"type": "command",
"command": "npx agentic-flow@alpha hooks pre-command \"$TOOL_INPUT_command\""
}]
}
]
}
```
### PostToolUse Hooks
Triggered after successful tool execution.
```json
{
"PostToolUse": [
{
"matcher": "Edit|Write|MultiEdit",
"hooks": [{
"type": "command",
"command": "npx agentic-flow@alpha hooks post-edit \"$TOOL_INPUT_file_path\" --success"
}]
}
]
}
```
### Session Hooks
Triggered at session start/end.
```json
{
"SessionStart": [{
"hooks": [{
"type": "command",
"command": "npx agentic-flow@alpha hooks intelligence stats --json"
}]
}],
"SessionEnd": [{
"hooks": [{
"type": "command",
"command": "npx agentic-flow@alpha hooks metrics --session --json"
}]
}]
}
```
## Hook Commands
```bash
# Pre-edit: Check for patterns, validate
npx agentic-flow@alpha hooks pre-edit "src/api.ts"
# Post-edit: Format, learn pattern
npx agentic-flow@alpha hooks post-edit "src/api.ts" --success
# Pre-command: Validate safety
npx agentic-flow@alpha hooks pre-command "npm install"
# Post-command: Track result
npx agentic-flow@alpha hooks post-command "npm install"
# Intelligence routing
npx agentic-flow@alpha hooks route "implement OAuth"
# Session metrics
npx agentic-flow@alpha hooks metrics --session
```
## Automation Features
### Auto-Agent Assignment
Hooks automatically assign specialized agents based on file type:
- `.ts`, `.tsx` → TypeScript agent
- `.py` → Python agent
- `.md` → Documentation agent
- `test.*` → Testing agent
### Command Validation
Pre-command hooks validate for:
- Dangerous patterns (rm -rf, force push)
- Permission requirements
- Resource limits
### Pattern Learning
Post-edit hooks learn:
- Code patterns
- Successful solutions
- Error recovery strategies
### Metrics Tracking
Session hooks track:
- Token usage
- Tool invocations
- Success rates
- Performance metrics
## Configuration
Full hook configuration in `.claude/settings.json`:
```json
{
"hooks": {
"PreToolUse": [...],
"PostToolUse": [...],
"PostToolUseFailure": [...],
"SessionStart": [...],
"SessionEnd": [...],
"UserPromptSubmit": [...],
"Stop": [...]
}
}
```
## Best Practices
1. **Keep hooks fast**: <100ms for interactive experience
2. **Use timeout**: Prevent hanging hooks
3. **Handle failures gracefully**: Use `|| true` for optional hooks
4. **Log metrics**: Track for optimization
5. **Test hooks**: Verify before deployment
Ver no GitHub