What is the maximum number of concurrent workers supported?

To prevent context overflow and maintain performance stability, the system is configured to support a maximum of 5 concurrent worker agents at any given time.

How do workers communicate with the main orchestrator?

Workers use a shared message bus to post heartbeats, status updates, and completion events, allowing the orchestrator to monitor progress and aggregate final results.

How does the skill handle task dependencies?

It parses a Directed Acyclic Graph (DAG) to identify 'parallel groups'—tasks at the same topological level that have no unmet dependencies, ensuring logical execution order.

Does this skill help with deadlocks?

Yes, it includes a deadlock detection and resolution protocol that monitors the message bus for cycles and automatically releases locks from the oldest worker to resume progress.

What happens if a worker agent fails during execution?

The system performs failure isolation, identifies which downstream tasks are blocked, and attempts automated recovery through a configurable retry strategy before escalating for manual intervention.

Parallel Multi-Agent Dispatcher

Name: Parallel Multi-Agent Dispatcher
Author: Ibrahim-3d

byIbrahim-3d

•

291

•

생산성 및 워크플로

Orchestrates simultaneous execution of multiple AI worker agents for complex, non-linear development tasks using DAG-based scheduling.

The Parallel Dispatch skill enables Claude Code to handle large-scale projects by breaking down Directed Acyclic Graphs (DAGs) into parallelizable tasks. It manages a pool of up to five ephemeral worker agents, providing robust monitoring, failure recovery, and deadlock resolution via a shared message bus. This is particularly useful for multi-file refactoring, concurrent testing, or complex architectural changes where independent tasks can be executed simultaneously without manual intervention, significantly reducing total execution time.

주요 기능

01Automated failure isolation and downstream impact analysis

02291 GitHub stars

03DAG-driven parallel execution for dependency-aware task scheduling

04Integrated deadlock detection and resolution strategies

05Concurrency management with a 5-worker pool to prevent context overflow

06Real-time monitoring using heartbeats and a shared message bus

사용 사례

01Executing batch documentation or styling updates across different repository directories

02Running a suite of integration tests in parallel while monitoring for specific worker failures

03Refactoring multiple independent modules across a large codebase simultaneously

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ibrahim-3d/conductor-orchestrator-superpowers parallel-dispatch

For use in Claude.ai and ChatGPT

Download Skill