Port Micro-Nuclear Energy Dispatch Engine with OpenAI Agents SDK
Major international seaports, such as the Port of Corpus Christi, are establishing micro-modular nuclear reactor partnerships to power shore power operations (cold ironing) and zero-emission container terminals. Efficient port microgrid operation requires dynamically balancing fluctuating nuclear power generation with unpredictable vessel shore power demand spikes. In this challenge, you will implement an energy dispatch agent using the OpenAI Agents SDK. The agent forecasts microgrid load, monitors SMR (Small Modular Reactor) thermal limits, and routes zero-emission electricity to docked container ships while mitigating grid surge tariffs.
What you are building
The core problem, expected build, and operating context for this challenge.
Build a port microgrid energy dispatch agent using OpenAI Agents SDK to balance nuclear generator output with ship shore power demands.
How work is evaluated
Evaluates load balancing performance and prevention of grid overload events.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
zero_blackout_guarantee
Total dispatched power must equal or exceed ship demand without exceeding SMR limits.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
clean_energy_utilization_pct
Percentage of total demand met via SMR and BESS without municipal grid pull • target: 100 • range: 0-100
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Implement OpenAI Agents SDK tool execution for microgrid dynamic dispatching
Model port electrical load demands based on vessel cold-ironing connections
Enforce micro-nuclear power plant safe operating constraints
Optimize real-time battery energy storage system (BESS) charging schedules
Reference links and supporting material
Port energy telemetry dataset containing hourly SMR power output, vessel cold-ironing loads, and spot grid electricity tariffs.
Challenge setup
Evaluates load balancing performance and prevention of grid overload events.
Challenge input
JSON with SMR megawatt capacity, battery charge state, and ship power MW demands
OpenAI Agents SDK
Required framework for building autonomous function-calling agents.
Evaluated output
JSON with MW dispatch values per asset and grid import balance
- Total dispatched power must equal or exceed ship demand without exceeding SMR limits.
- Percentage of total demand met via SMR and BESS without municipal grid pull • target: 100 • range: 0-100
- Benchmark: EnergyBench-Grid
- Clean Energy Utilization Pct target: 100
- 1 public reference case
- 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
- OpenAI · Required
- CleverBee · Optional
- Cekura · Optional
Evaluation contract
- zero_blackout_guarantee · Weight 1
- clean_energy_utilization_pct · Weight 1
Metrics
- clean_energy_utilization_pct · Target 100
Benchmarks
- EnergyBench-Grid
- energy
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON with SMR megawatt capacity, battery charge state, and ship power MW demands
Output: JSON with MW dispatch values per asset and grid import balance
Execution harness: Python
Evaluation tools
- OpenAI Agents SDK · Agent Framework
Recipe freeze marker
No recipe freeze time is recorded.