Skip to content
Tresca

CMG AI integrations

No MCP server or agent skill for CMG is listed yet (checked 4 Nov 2026). No MCP server or agent skill exists for CMG's simulators (IMEX, GEM, STARS) or CMOST. CMG's own AI so far is machine learning inside CMOST for history matching, optimisation and proxy models; on its August 2026 earnings call management described AI agents for reservoir engineers as working prototypes. A community integration would work from CMG's keyword datasets and result files.

CMGNone listed yetChecked 4 Nov 2026

How AI connects to CMG

An MCP server or skill can only do what CMG lets programs do. CMG exposes 2 automation interfaces:

  • Simulator datasets (IMEX, GEM, STARS) (Input files): IMEX (black oil and unconventional), GEM (compositional) and STARS (thermal and advanced processes) are run from text datasets that Builder, CMG's pre-processor, writes and that engineers often edit by hand. Results are viewed in CMG Results. Text datasets are the most realistic entry point for an agent: it can template cases, change keywords and submit runs without a GUI. Documentation
  • CMOST (Other): CMOST varies dozens to hundreds of dataset parameters at once for history matching, optimisation and uncertainty work, using proxy models, machine learning and proxy-based acceptance-rejection sampling, and can send runs to the cloud. An agent sitting above CMOST would set up the parameter space and read back the ranked results rather than run each case itself. Documentation

What CMG offers

CMG's website markets CMOST's "Augmented Intelligence" history matching and machine-learning proxies, and a customer quote mentions "CMOST AI". A transcript of CMG's first-quarter fiscal 2027 earnings call (12 August 2026) quotes management building a common architecture for AI agents across its simulators, with prototypes that launch and monitor runs. No product release or MCP server had been announced as of 4 November 2026.

Checked 4 Nov 2026. Sources: CMOST (CMG), CMG home page, CMG Q1 fiscal 2027 earnings call transcript (roic.ai)

What an integration for CMG could do

  • Build IMEX or GEM sensitivity datasets from a base case by changing permeability multipliers, well constraints or injection schedules.
  • Check a dataset for missing include files and inconsistent units before a long thermal STARS run.
  • Summarise a set of finished runs, comparing cumulative oil, water cut and steam-oil ratio across cases.
  • Set up a CMOST study by proposing parameters and ranges from the dataset and the objective you describe.

Know of one, or building one?

If an MCP server or agent skill for CMG exists, tell us and we'll check it and list it. You can also ask for one, which helps builders see demand.

Submit a listing Request an integration

Tools with AI integrations for similar work

OPM FlowOpen source1 serverECLIPSE / INTERSECTSLB1 servertNavigatorRock Flow Dynamics1 serverpyResToolboxOpen source2 servers · 1 skill

Questions

Is there an MCP server for CMG IMEX, GEM or STARS?

No. None was found on 4 November 2026, and CMG hasn't announced one. Its AI work is in CMOST and in agent prototypes it mentioned to investors.

Can an AI agent edit CMG datasets?

Any coding agent can edit the text datasets, but no tool on Tresca knows CMG keywords specifically. For ECLIPSE-format work, opm-mcp shows what a deck-aware server looks like.

AI-generated models, drawings and results must be checked by a qualified engineer before use.

Sources (checked 4 Nov 2026): CMG home page · IMEX (CMG) · CMOST (CMG) · CMG product page

CMG is a trademark of CMG. Tresca is independent and not affiliated with or endorsed by CMG.