Files
SuperClaude/.claude/commands/spawn.md
NomenAK d24503ca02 refactor: Complete SuperClaude v2 migration with @include reference system
- Migrate all command files to use @include reference system
- Consolidate shared patterns into new yml structure
- Create central superclaude shared configuration files
- Remove deprecated markdown files (MCP.md, PERSONAS.md, RULES.md)
- Add new documentation structure in docs/
- Update installation script for new architecture
- Add ROADMAP.md and VERSION files

This completes the major architectural refactor to improve maintainability
and reduce duplication across the SuperClaude command system.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-06-25 16:51:53 +02:00

140 lines
2.9 KiB
Markdown

**Purpose**: Parallel specialized agents
---
@include shared/universal-constants.yml#Universal_Legend
## Command Execution
Execute: immediate. --plan→show plan first
Legend: Generated based on symbols used in command
Purpose: "[Action][Subject] in $ARGUMENTS"
Spawn specialized sub-agents for focused tasks with parallel execution capabilities.
@include shared/flag-inheritance.yml#Universal_Always
Examples:
- `/spawn --agent researcher "OAuth 2.0 best practices"` - Research then implement
- `/spawn --mode parallel --agent builder "User auth, Profile API"` - Parallel development
- `/spawn --mode sequential "Research → Build → Review payment"` - Full cycle workflow
- `/spawn --mode collaborative --ultrathink "Design microservices"` - Collaborative design
## Agent Types
Researcher Agent:
- Deep dive into topics
- Compare solutions
- Analyze trade-offs
- Find best practices
- Document findings
Builder Agent:
- Generate code
- Implement features
- Create tests
- Build prototypes
- Integrate systems
Reviewer Agent:
- Code quality checks
- Security analysis
- Performance review
- Best practice validation
- Suggest improvements
Optimizer Agent:
- Performance profiling
- Resource optimization
- Algorithm improvements
- Database tuning
- Cache strategies
Documenter Agent:
- API documentation
- User guides
- Code comments
- Architecture docs
- README files
## Execution Modes
Sequential Mode:
```yaml
Flow: Agent1 → Agent2 → Agent3
Use: When tasks depend on each other
Example: Research → Build → Review
```
Parallel Mode:
```yaml
Flow: Agent1 | Agent2 | Agent3
Use: For independent tasks
Example: Multiple feature builds
```
Collaborative Mode:
```yaml
Flow: Agents work together
Use: Complex problems
Example: System design session
```
## Best Practices
Task Definition:
- Clear objectives
- Specific deliverables
- Success criteria
- Resource limits
- Time constraints
Agent Selection:
- Match expertise to task
- Consider dependencies
- Plan coordination
- Set boundaries
- Define handoffs
Coordination:
- Clear communication
- Shared context
- Progress tracking
- Result integration
- Quality control
## Examples
```bash
# Research then implement
/spawn --agent researcher "OAuth 2.0 best practices"
/spawn --agent builder "Implement OAuth based on research"
# Parallel feature development
/spawn --mode parallel --agent builder "User auth, Profile API, Settings UI"
# Full cycle with review
/spawn --mode sequential "Research → Build → Review payment integration"
# Collaborative system design
/spawn --mode collaborative --ultrathink "Design microservices architecture"
```
## Integration
Works with:
- All command flags pass through
- Inherits persona preferences
- Shares project context
- Accumulates findings
- Coordinates outputs
## Deliverables
- Agent execution logs
- Task completion reports
- Integrated results
- Performance metrics
- Lessons learned
- Handoff documentation
@include shared/universal-constants.yml#Standard_Messages_Templates