How does parallel execution reduce execution time?

By running multiple independent tasks simultaneously instead of one after another, it reduces the total execution time from the sum of all tasks to approximately the duration of the single longest task.

When should I avoid using parallel execution?

Avoid parallelization when tasks have dependencies (Task B needs Task A's output), when a strict order is required (e.g., commit then push), or when multiple tasks need to modify the same file.

How do I retrieve results from parallel tasks?

After launching tasks with run_in_background: true, wait for completion and then use the TaskOutput tool with the corresponding task IDs to retrieve the individual results for synthesis.

What is the most critical rule for ensuring tasks run in parallel?

All Task tool calls must be included within a single assistant message. If Task calls are placed in separate messages, Claude Code will execute them sequentially.

Parallel Execution Patterns

Name: Parallel Execution Patterns
Author: Tahir-yamin

byTahir-yamin

•

Productividad y Flujo de Trabajo

Orchestrates multiple independent subagents simultaneously to accelerate complex development tasks and reduce execution time.

Acerca de

This skill provides standardized patterns for implementing parallel execution in Claude Code, allowing developers to spawn multiple subagents concurrently via the Task tool. By utilizing the run_in_background flag and consolidating tool calls into a single message, it significantly reduces total execution time for independent tasks such as multi-file analysis, perspective-based reviews, and modular feature implementation. It includes specific protocols for identifying parallelizable work, dynamic prompting for specialized roles, and effective result synthesis.

Características Principales

Integrated TodoWrite status tracking for parallel workflows
3 GitHub stars
Result synthesis patterns for unifying multi-agent outputs
Dynamic subagent prompting for specialized domain roles
Single-message execution protocols for maximum parallelism
Concurrent subagent spawning using background task orchestration

Casos de Uso

Performing multi-perspective reviews (Security, Performance, Architecture) simultaneously
Implementing independent modules or API endpoints in parallel to save time
Analyzing large codebases by scanning multiple directories concurrently

Acerca de

Características Principales

Integrated TodoWrite status tracking for parallel workflows
3 GitHub stars
Result synthesis patterns for unifying multi-agent outputs
Dynamic subagent prompting for specialized domain roles
Single-message execution protocols for maximum parallelism
Concurrent subagent spawning using background task orchestration

Casos de Uso

Performing multi-perspective reviews (Security, Performance, Architecture) simultaneously
Implementing independent modules or API endpoints in parallel to save time
Analyzing large codebases by scanning multiple directories concurrently