skills.rest
jakechen1
jakechen1COMMUNITY
Explore all skills by this author →
WebsiteGitHub

task-router

Estimate computational task costs and route them to the cheapest compatible compute tier.

Updated Apr 9, 2026
Search Tags:#automation#gpu#cost-estimation#task-routing#tier-selection#local-compute#cheaha
One-click install
npx skills add https://github.com/jakechen1/echo-research-framework --skill task-router-jakechen1
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: task-router
Source: https://github.com/jakechen1/echo-research-framework/tree/main/skills/task-router
Command: npx skills add https://github.com/jakechen1/echo-research-framework --skill task-router-jakechen1

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, python3, ssh, and includes scripts (resource) components.

What problem does it solve?

Estimate and route computational tasks to the cheapest compatible compute tier, reducing waste and cost while preserving performance.

Core Features & Use Cases

  • Cost-aware routing: estimates CPU time (seconds), memory (RAM) and GPU needs to select the cheapest viable tier (local serial, local parallel-10, W0/L0 local, or Cheaha HPC).
  • Tier-aware orchestration: supports local execution, parallelization, and remote Cheaha submission for large workloads.
  • Telemetry-driven calibration and repeatable decisions across multiple tasks.

Quick Start

Call the route.sh script with your task name, item count, and operation type to receive the recommended tier.

Frequently Asked Questions about task-router

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

FAQPage Schema
How do I route computational tasks to the cheapest compute tier?▼

Cost-aware task routing directs workloads to the cheapest viable compute tier by estimating CPU time, RAM, and GPU needs. It evaluates local serial, local parallel, and Cheaha HPC options to minimize compute waste.

How do I estimate compute costs for GPU and CPU workloads before submission?▼

Compute cost estimation uses per-task telemetry and a cost table to project resource requirements. It analyzes item count and operation type to estimate CPU seconds and RAM needed for tier selection.

Does task routing work with Cheaha HPC for large compute jobs?▼

Yes, task routing supports remote Cheaha HPC submission for large workloads. It evaluates GPU or CPU requirements and can automatically submit jobs to Cheaha when local compute tiers lack sufficient capacity.

What is the best way to automate local compute tier selection for parallel tasks?▼

Automating local compute tier selection involves calling a routing script with task name, item count, and operation type. The script analyzes telemetry to recommend local serial or local parallel execution based on available resources.

Do I need ssh and python3 installed to use automated compute tier routing?▼

Yes, automated compute tier routing requires ssh, python3, and jq installed. These dependencies enable the routing script to process telemetry data and submit jobs to remote Cheaha HPC clusters via SSH.

Related Skills

Explore More Software Engineering Skills→
hve4638hve4638

serverless-modal

Automate GPU workload deployment and orchestration on Modal's serverless platform.

Community
Advanced
jakechen1jakechen1

cheaha-hpc

Submit, monitor, and manage Slurm jobs on the Cheaha HPC cluster.

Community
Advanced
plantcadplantcad
38

tacc

Automate experiment execution on the TACC Vista HPC cluster.

Official
Advanced
michaelrizvimichaelrizvi
11

slurm

Write, debug, and optimize SLURM job scripts for HPC clusters.

Community
Intermediate