How does it handle dependencies between tasks?

The skill creates a master plan to map dependencies and executes tasks in topological order, ensuring all prerequisites are successfully completed before starting dependent sub-tasks.

Can I track the progress of long-running operations?

Yes, the orchestrator maintains a .orchestrator/master_plan.md file that provides a visual status (Pending, Running, Done, Failed) for every sub-task in the workflow.

Do I need to manually launch sub-agents?

By default, the skill uses simulated parallel execution within the current session. If requested, it can integrate with the Claude CLI to launch real, distributed sub-agents for true parallel processing.

What happens if a sub-task fails during execution?

The orchestrator implements robust error handling, offering strategy options like three-step retries with exponential backoff, skipping non-critical tasks, or a fail-fast approach for critical dependencies.

Distributed Task Orchestrator

Name: Distributed Task Orchestrator
Author: xuzeyu91

byxuzeyu91

•

生产力与工作流

Decomposes complex multi-step requests into atomic sub-tasks for parallel execution and automated result aggregation.

The Distributed Task Orchestrator empowers Claude to act as a sophisticated project manager for high-complexity coding tasks. By breaking down large requests into independent atomic sub-tasks, it maps dependencies and manages execution through either simulated parallel batches or real Claude CLI sub-agents. This skill is essential for large-scale operations such as codebase refactoring, multi-file migrations, and batch data processing, providing a structured workflow with a clear audit trail through a dedicated orchestration directory.

主要功能

01Dynamic task decomposition with dependency mapping and DAG support

02Real-time progress tracking via a centralized master_plan.md log

0313 GitHub stars

04Automated result aggregation and merging based on task dependency order

05Parallel execution via simulated batches or real Claude CLI sub-agent launching

06Configurable error handling including retries and fail-fast strategies

使用场景

01Executing complex migration plans that require strict multi-step dependency management

02Performing batch processing on a large set of data or configuration files

03Refactoring a large codebase across multiple directories simultaneously

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add xuzeyu91/ai-agent-toolkit distributed-task-orchestrator

For use in Claude.ai and ChatGPT

Download Skill