Skip to main content
Back to challenges
Challenge

Build a Data Center Thermal Load Balancing Scheduler with Claude Agents SDK

Construct an intelligent compute workload scheduler for Singapore's 200MW data center infrastructure. Leveraging Claude Agents SDK with extended thinking and computer use capabilities, your agent will balance server power allocation against green energy constraints and thermal limits.

Workflow AutomationHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Build an intelligent load-balancing agent using Claude Agents SDK to manage thermal limits and energy targets in Singapore data centers.

Delivery guide

How work is evaluated

Evaluation

Evaluate capacity planning assignments for compliance with 200MW grid and PUE constraints.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

thermal_limit_safety

No hall temp exceeds maximum safety limit of 28C

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

pue_score

Calculated Power Usage Effectiveness • target: 1.22 • range: 1-2

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Utilize Claude Agents SDK with extended thinking mode for complex constraint optimization

  • Design multi-objective scheduling tool definitions for rack power and thermal management

  • Enforce maximum Power Usage Effectiveness (PUE) target thresholds

  • Automate workload shifting based on dynamic green grid availability signals

Resources and assets

Reference links and supporting material

Dataset notes

Telemetry logs from simulated 200MW data center facility representing hall temperatures, PDU loads, and cooling energy.

Challenge setup

Evaluate capacity planning assignments for compliance with 200MW grid and PUE constraints.

Challenge input

JSON containing hall_states (temp, load_mw) and incoming_job_mw

Claude Agents SDK

Anthropic SDK supporting extended thinking capabilities.

Evaluated output

JSON with assigned_hall, revised_hall_temps, and pue_estimate

Evaluation checks
  • No hall temp exceeds maximum safety limit of 28C
  • Calculated Power Usage Effectiveness • target: 1.22 • range: 1-2
Configured evaluation evidence
  • Benchmark: DC_Energy_Opt_Bench
  • Pue Score target: 1.22
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • RAI · Required
  • Agentic Dev · Optional
  • AutoGluon · Optional

Evaluation contract

  • thermal_limit_safety · Weight 1
  • pue_score · Weight 1

Metrics

  • pue_score · Target 1.22

Benchmarks

  • DC_Energy_Opt_Bench
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing hall_states (temp, load_mw) and incoming_job_mw

    Output: JSON with assigned_hall, revised_hall_temps, and pue_estimate

    Execution harness: Python

Evaluation tools

  • Claude Agents SDK · Agent SDK

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a Data Center Thermal Load Balancing Scheduler with Claude Agents SDK