Parallel Execution

Parallel Execution Patterns

When to Load

• Trigger: Multi-agent tasks, concurrent operations, spawning subagents, parallelizing independent work
• Skip: Single-step tasks or sequential workflows with no parallelization opportunity

Core Concept

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.

Execution Protocol

Step 1: Identify Parallelizable Tasks

Before spawning, verify tasks are independent:

• No task depends on another's output
• Tasks target different files or concerns
• Can run simultaneously without conflicts

Step 2: Prepare Dynamic Subagent Prompts

Each subagent receives a custom prompt defining its role:

Copy code
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]

Step 3: Launch All Tasks in ONE Message

CRITICAL: Make ALL Task calls in the SAME assistant message:

Copy code
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

Step 4: Retrieve Results with TaskOutput

After launching, retrieve each result:

Copy code
[Wait for completion, then retrieve]

TaskOutput: task_1_id
TaskOutput: task_2_id
TaskOutput: task_3_id

Step 5: Synthesize Results

Combine all subagent outputs into unified result:

• Merge related findings
• Resolve conflicts between recommendations
• Prioritize by severity/importance
• Create actionable summary

Dynamic Subagent Patterns

Pattern 1: Task-Based Parallelization

When you have N tasks to implement, spawn N subagents:

Copy code
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

Pattern 2: Directory-Based Parallelization

Analyze multiple directories simultaneously:

Copy code
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

Pattern 3: Perspective-Based Parallelization

Review from multiple angles simultaneously:

Copy code
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

TodoWrite Integration

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

Copy code
# 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" }
]

When to Use Parallel Execution

Good candidates:

• Multiple independent analyses (code review, security, tests)
• Multi-file processing where files are independent
• Exploratory tasks with different perspectives
• Verification tasks with different checks
• Feature implementation with independent components

Avoid parallelization when:

• Tasks have dependencies (Task B needs Task A's output)
• Sequential workflows are required (commit -> push -> PR)
• Tasks modify the same files (risk of conflicts)
• Order matters for correctness

Performance Benefits

Parallel execution is approximately Nx faster where N is the number of independent tasks.

Example: Feature Implementation

User request: "Implement user authentication with login, registration, and password reset"

Orchestrator creates plan:

1. Implement login endpoint
2. Implement registration endpoint
3. Implement password reset endpoint
4. Add authentication middleware
5. Write integration tests

Parallel execution:

Copy code
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]

Troubleshooting

Tasks running sequentially?

• Verify ALL Task calls are in SINGLE message
• Check run_in_background: true is set for each

Results not available?

• Use TaskOutput with correct task IDs
• Wait for tasks to complete before retrieving

Conflicts in output?

• Ensure tasks don't modify same files
• Add conflict resolution in synthesis step