Patterns for parallel subagent execution using Task tool with run_in_background. Use when coordinating multiple independent tasks, spawning dynamic subagents, or implementing features that can be parallelized.
Spawns multiple subagents simultaneously using the Task tool with `run_in_background: true`. Use when you have independent tasks that can run concurrently, like analyzing different directories or implementing parallel features. All Task calls must be in a single message for true parallelism.
/plugin marketplace add CloudAI-X/claude-workflow/plugin install project-starter@claude-workflowThis skill inherits all available tools. When active, it can use any tool Claude has access to.
Parallel execution spawns multiple subagents simultaneously using the Task tool with run_in_background: true. This enables N tasks to run concurrently, dramatically reducing total execution time.
Critical Rule: ALL Task calls MUST be in a SINGLE assistant message for true parallelism. If Task calls are in separate messages, they run sequentially.
Before spawning, verify tasks are independent:
Each subagent receives a custom prompt defining its role:
You are a [ROLE] specialist for this specific task.
Task: [CLEAR DESCRIPTION]
Context:
[RELEVANT CONTEXT ABOUT THE CODEBASE/PROJECT]
Files to work with:
[SPECIFIC FILES OR PATTERNS]
Output format:
[EXPECTED OUTPUT STRUCTURE]
Focus areas:
- [PRIORITY 1]
- [PRIORITY 2]
CRITICAL: Make ALL Task calls in the SAME assistant message:
I'm launching N parallel subagents:
[Task 1]
description: "Subagent A - [brief purpose]"
prompt: "[detailed instructions for subagent A]"
run_in_background: true
[Task 2]
description: "Subagent B - [brief purpose]"
prompt: "[detailed instructions for subagent B]"
run_in_background: true
[Task 3]
description: "Subagent C - [brief purpose]"
prompt: "[detailed instructions for subagent C]"
run_in_background: true
After launching, retrieve each result:
[Wait for completion, then retrieve]
TaskOutput: task_1_id
TaskOutput: task_2_id
TaskOutput: task_3_id
Combine all subagent outputs into unified result:
When you have N tasks to implement, spawn N subagents:
Plan:
1. Implement auth module
2. Create API endpoints
3. Add database schema
4. Write unit tests
5. Update documentation
Spawn 5 subagents (one per task):
- Subagent 1: Implements auth module
- Subagent 2: Creates API endpoints
- Subagent 3: Adds database schema
- Subagent 4: Writes unit tests
- Subagent 5: Updates documentation
Analyze multiple directories simultaneously:
Directories: src/auth, src/api, src/db
Spawn 3 subagents:
- Subagent 1: Analyzes src/auth
- Subagent 2: Analyzes src/api
- Subagent 3: Analyzes src/db
Review from multiple angles simultaneously:
Perspectives: Security, Performance, Testing, Architecture
Spawn 4 subagents:
- Subagent 1: Security review
- Subagent 2: Performance analysis
- Subagent 3: Test coverage review
- Subagent 4: Architecture assessment
When using parallel execution, TodoWrite behavior differs:
Sequential execution: Only ONE task in_progress at a time
Parallel execution: MULTIPLE tasks can be in_progress simultaneously
# Before launching parallel tasks
todos = [
{ content: "Task A", status: "in_progress" },
{ content: "Task B", status: "in_progress" },
{ content: "Task C", status: "in_progress" },
{ content: "Synthesize results", status: "pending" }
]
# After each TaskOutput retrieval, mark as completed
todos = [
{ content: "Task A", status: "completed" },
{ content: "Task B", status: "completed" },
{ content: "Task C", status: "completed" },
{ content: "Synthesize results", status: "in_progress" }
]
Good candidates:
Avoid parallelization when:
| Approach | 5 Tasks @ 30s each | Total Time |
|---|---|---|
| Sequential | 30s + 30s + 30s + 30s + 30s | ~150s |
| Parallel | All 5 run simultaneously | ~30s |
Parallel execution is approximately Nx faster where N is the number of independent tasks.
User request: "Implement user authentication with login, registration, and password reset"
Orchestrator creates plan:
Parallel execution:
Launching 5 subagents in parallel:
[Task 1] Login endpoint implementation
[Task 2] Registration endpoint implementation
[Task 3] Password reset endpoint implementation
[Task 4] Auth middleware implementation
[Task 5] Integration test writing
All tasks run simultaneously...
[Collect results via TaskOutput]
[Synthesize into cohesive implementation]
Tasks running sequentially?
run_in_background: true is set for eachResults not available?
Conflicts in output?
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.