scheduler

Coordinate resource allocation and admission control for Codegg daemon jobs.

Updated May 6, 2026

Search Tags:#concurrency #rust #daemon #admission-control #scheduler #fair-queue

Ready to run

npx skills add https://github.com/dbowm91/codegg --skill scheduler-dbowm91

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: scheduler
Source: https://github.com/dbowm91/codegg/tree/main/.opencode/skills/scheduler
Command: npx skills add https://github.com/dbowm91/codegg --skill scheduler-dbowm91

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill manages the global admission control for the Codegg daemon, preventing resource exhaustion and ensuring fair execution priority across multiple concurrent agent tasks.

Core Features & Use Cases

Fair Queueing: Implements priority classes, lanes, and aging to prevent task starvation.
Admission Control: Provides atomic permit acquisition for CPU, memory, and I/O resources.
Typed Execution: Supports diverse job executors including test runners, subagents, and managed processes.
Use Case: When running multiple background agents, the scheduler ensures that high-priority interactive tasks are not blocked by long-running maintenance jobs.

Quick Start

Use the scheduler skill to verify the current admission status and resource permit availability for a new job submission.

Frequently Asked Questions about scheduler

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

FAQPage Schema

How does fair queueing prevent task starvation in a concurrent agent environment?▼

Fair queueing prevents task starvation by implementing priority classes, lanes, and aging. This ensures lower-priority background jobs eventually acquire execution permits, even when high-priority interactive tasks continuously compete for resources.

How do I manage admission control for concurrent background agents?▼

You manage admission control for concurrent agents by acquiring atomic resource permits for CPU, memory, and I/O. A centralized reconciliation loop coordinates allocation between durable job stores and typed executors to enforce system-wide policies.

Can I configure priority-based resource budgets for different job executors?▼

Yes, you can configure priority-based resource budgets for diverse job executors, including test runners and subagents. The scheduler enforces these budgets through a centralized reconciliation loop and process-spawn provenance.

What is the best way to ensure high-priority interactive tasks are not blocked by maintenance jobs?▼

The best way to prevent blocking is using a scheduler with priority classes and atomic permit acquisition. This ensures high-priority interactive tasks acquire CPU and memory permits before long-running maintenance jobs.

Why does my concurrent agent task execution experience resource exhaustion?▼

Resource exhaustion occurs when concurrent tasks lack global admission control. Without atomic permit acquisition and fair queueing, unmanaged processes overwhelm system CPU, memory, and I/O budgets.

Does the scheduler support durable job stores for managed process execution?▼

Yes, the scheduler supports durable job stores for managed process execution. It coordinates resource allocation between these stores and typed executors using fair-queueing and priority-based permit acquisition.