Jafurah Gas Expansion Risk Workflow with LangChain & LangGraph
A frontrunner emerged for the Jafurah fifth expansion phase in Saudi Arabia. Build a stateful LangGraph agent that models wellhead drilling expansion workflows, monitors regulatory approvals, and predicts environmental compliance risks. Maintain state persistence and achieve 100% compliance audit traceability.
What you are building
The core problem, expected build, and operating context for this challenge.
Construct a stateful LangGraph state machine with LangChain tools to automate risk checking for the Jafurah gas field expansion project.
How work is evaluated
Evaluates graph transition flow, environmental risk detection, and audit trail logging.
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.
human_in_loop_trigger
Ensures high-risk sites correctly halt for human intervention
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
workflow_completion_rate
Percentage of test inputs processed without node failure • target: 1 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Build a stateful directed graph using LangGraph for multi-stage gas project risk analysis.
Implement persistent memory checkpointers in LangGraph.
Parse unconventional gas extraction permits and environmental impact standard documents.
Enforce human-in-the-loop review nodes for high-risk environmental waivers.
Reference links and supporting material
Environmental baseline data and wellsite engineering specs for Saudi Arabia's Jafurah basin.
Challenge setup
Evaluates graph transition flow, environmental risk detection, and audit trail logging.
Challenge input
JSON containing well site specs, groundwater proximity, and flare estimates
LangChain
Provides prompt management and tool wrappers.
LangGraph
Enables stateful agent graph flows with memory.
Evaluated output
JSON containing risk classification, required permits, and state transition history
- Ensures high-risk sites correctly halt for human intervention
- Percentage of test inputs processed without node failure • target: 1 • range: 0-1
- Benchmark: WorkflowGraphBench
- Workflow Completion Rate 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
- LangGraph · Required
- Langchain · Optional
- Merge · Optional
- LangGraph · Required
- Langchain · Optional
Evaluation contract
- human_in_loop_trigger · Weight 1
- workflow_completion_rate · Weight 1
Metrics
- workflow_completion_rate · Target 1
Benchmarks
- WorkflowGraphBench
- energy
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON containing well site specs, groundwater proximity, and flare estimates
Output: JSON containing risk classification, required permits, and state transition history
Execution harness: Python
Evaluation tools
- LangChain · LLM Integration
- LangGraph · Stateful Orchestration
Recipe freeze marker
No recipe freeze time is recorded.