mesh

Delegate inline task lists to subagents with dependency-aware parallel execution.

66|2|Updated Feb 18, 2015
One-click install
npx skills add https://github.com/tkersey/dotfiles --skill mesh
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mesh
Source: https://github.com/tkersey/dotfiles/tree/main/codex/skills/mesh
Command: npx skills add https://github.com/tkersey/dotfiles --skill mesh

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill addresses the challenge of coordinating multiple tasks by delegating work to spawned subagents, enabling parallel execution while preserving correct ordering through dependency-aware waves.

Core Features & Use Cases

  • Inline parallel task execution: parse and schedule inline task lists, spawn workers, and manage frontier execution.
  • Dependency-respecting waves: ensure tasks run in an order that honors dependencies and commits/validates per step.
  • Safe orchestration: orchestrator handles planning, liveness monitoring, and post-step validation; avoids direct task execution unless explicitly allowed.
  • Use cases include large task batches, patch/apply workflows, and automated multi-step operations that require worker tooling.

Quick Start

Invoke the mesh skill with an inline task list using the $mesh command in your message. For example, provide a list of tasks; then reply with $mesh confirm to run exactly this task list (no copy/paste).

Frequently Asked Questions about mesh

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run parallel tasks with dependencies using subagents?

To execute parallel tasks with dependency management, provide an inline task list and invoke the mesh orchestrator, which spawns worker subagents to execute tasks in dependency-respecting waves. The orchestrator schedules tasks, monitors liveness, and validates post-step results while workers handle implementation.

What is dependency-aware parallel execution for multi-step workflows?

Dependency-aware parallel execution is an orchestration mechanism that groups tasks into sequential waves, ensuring dependent tasks only run after their prerequisites complete. It coordinates spawned worker subagents to handle parallel workloads while preserving correct execution order and validating each step.

How do I start an automated multi-step workflow with an inline task list?

To start an automated multi-step workflow, invoke the mesh skill with an inline task list using the $mesh command in your message. Reply with $mesh confirm to execute that exact task list without copy/pasting, allowing the orchestrator to delegate implementation work to spawned worker subagents.

Can the orchestrator execute tasks directly instead of spawning subagents?

No, the orchestrator is forbidden from executing tasks directly unless explicitly authorized by the user. It handles planning, liveness monitoring, and post-step validation, while delegating actual task implementation to spawned worker subagents to maintain safe orchestration boundaries.

When should I use parallel worker subagents for task automation?

Use parallel worker subagents for large task batches, patch/apply workflows, and automated multi-step operations that require worker tooling. It is ideal when you have an explicit task list that benefits from dependency-respecting parallel execution with per-task commits and post-step validation.

Does parallel task orchestration handle per-task commits and validation?

Yes, dependency-respecting parallel execution ensures tasks run in an order that honors dependencies and performs commits and validation per step. The orchestrator manages post-step validation after each worker subagent completes its assigned task within the execution wave.