Agent Orchestration
Available Agents
Located in .claude/agents/:
| Agent | Purpose | When to Use |
|---|---|---|
| planner | Implementation planning | Complex features, refactoring |
| architect | System design | Architectural decisions |
| tdd-guide | Test-driven development | New features, bug fixes |
| code-reviewer-backend | Backend code review | After writing Ktor/server-side code |
| code-reviewer-domain | Domain layer code review | After writing domain models, use cases, repositories |
| code-reviewer-frontend | Frontend code review | After writing UI, ViewModels, Compose screens |
| build-resolver | Fix build errors | When build or compilation fails |
| doc-updater | Documentation (technical, functional, articles) | After features, business rule changes, or publishing |
Immediate Agent Usage
No user prompt needed — launch proactively when the situation calls for it:
- Complex feature request → Use planner agent first
- Architectural decision → Use architect agent
- Bug fix or new feature → Use tdd-guide agent
- Code just written or modified → Use the appropriate code-reviewer agent
- Build or compilation fails → Use build-resolver agent
- Feature shipped or business rules updated → Use doc-updater agent
Choosing the Right Code Reviewer
There are three specialized code reviewers — pick based on what changed:
| Changed area | Use |
|---|---|
| Ktor routes, DTOs, API plugins, server config | code-reviewer-backend |
| Domain models, use cases, repository interfaces, business logic | code-reviewer-domain |
| Compose UI, ViewModels, navigation, platform UI code | code-reviewer-frontend |
When a change spans multiple layers, launch all relevant reviewers in parallel.
Parallel Task Execution
Always use parallel Task execution for independent operations:
# GOOD: Parallel execution
Launch 3 agents in parallel:
1. Agent 1: Domain review of new Task entity
2. Agent 2: Backend review of new Task endpoint
3. Agent 3: Frontend review of new Task screen
# BAD: Sequential when unnecessary
First agent 1, then agent 2, then agent 3
Multi-Perspective Analysis
For complex problems, use split-role sub-agents to get independent perspectives:
- Domain correctness reviewer
- Senior engineer (architecture and design)
- Security expert
- Consistency reviewer (naming, patterns, conventions)
- Test coverage reviewer