simulation_skill

You can automate OPM Flow reservoir simulations by directing an agent to execute a DATA file, which manages the run lifecycle, configures MPI and threads, and handles output directories for deterministic execution.

Monitoring for background reservoir simulation runs works by parsing PRT file tails and tracking status, with optional LLM-assisted summaries to provide progress updates and detect failures across foreground and background executions.

Yes, you can use LangChain agents to auto-fix failed OPM Flow runs through ReAct-based integration that automatically detects errors, parses reports, and heals inputs within human-in-the-loop or direct execution workflows.

The best way to manage simulation run lifecycles is through deterministic tool execution that supports stopping, pausing, and healing runs while tracking logs and metadata across human-in-the-loop confirmations and scenario test chains.

Yes, you need Pydantic and LangChain dependencies along with an LLM provider and PyYAML to enable agent-driven automation, structured data validation, and optional auto-fixing capabilities for reservoir simulation execution.