MCP serverOfficial
tNavigator Local MCP Serverby Rock Flow Dynamics
The MCP server built into tNavigator 26.1 and later, letting third-party AI apps search its docs and manage projects.
tNavigator MCP servers and AI agent skills
tNavigator is one of the few reservoir tools with a vendor MCP server: since version 26.1 its AI Assistant includes a Local MCP Server that lets third-party AI applications search tNavigator's documentation and manage projects through its API. Rock Flow Dynamics hasn't published the tool list or setup steps publicly. No community MCP server or skill was found.
- Includes Rock Flow Dynamics's official server.
- 1 MCP server listed.
Rock Flow Dynamics1 serverChecked 4 Nov 2026
Integrations for tNavigator
Official
Listings are sorted by origin, then by recent activity. GitHub data is refreshed nightly. Nothing here is ranked.
How AI connects to tNavigator
An MCP server or skill can only do what tNavigator lets programs do. tNavigator exposes 2 automation interfaces:
- tNavigator Python API (Python): Rock Flow Dynamics describes a Python API for automating workflows, editing simulation models, running simulations and extracting results, and for bringing open-source machine-learning libraries into tNavigator workflows. Python code also runs inside workflow Custom Code steps, which the AI Assistant can write or edit on request. The full API reference sits behind the client support portal rather than on public pages. Documentation
- Control from Petex RESOLVE (OpenServer wrapper) (COM automation): Petex documents a generic OpenServer wrapper that lets RESOLVE control third-party applications, naming tNavigator alongside Eclipse and HYSYS. That gives integrated production models a way to run tNavigator cases with the same DoSet, DoCmd and DoGet commands used for PROSPER and GAP, which an agent working through OpenServer could reuse. Documentation
What Rock Flow Dynamics offers
tNavigator's AI Assistant is a multi-agent assistant inside the interface, running by default on a local server inside your network and connecting to LLM servers over the OpenAI API protocol. It answers questions, helps set up and troubleshoot models, writes Python workflow code, and its AI Summary explains simulation log warnings. Version 26.1 added a Local MCP Server, and 26.3 upgraded the assistant. A webinar on 18 November covers MCP and reusable skills.
Checked 4 Nov 2026. Sources: tNavigator AI Assistant (Rock Flow Dynamics), tNavigator 26.3 key features (Rock Flow Dynamics), Live webinar: Agentic AI & MCP in tNavigator (Rock Flow Dynamics)
What you can do with tNavigator and AI
- Search tNavigator's documentation from Claude or another MCP client while you build a model, instead of switching to the help system.
- Manage tNavigator projects from an external AI client through the Local MCP Server's API functions, for example to find and open the case you want to review.
- Ask the built-in AI Summary why a run failed, with the warnings and errors from the calculation log explained and fixes suggested.
- Generate or edit Python workflow code for a repetitive task, such as exporting well results for every case in a project.
Taken from the listings' own documentation. Check each listing for exactly what it supports.
Related tools
Questions
Does tNavigator have an MCP server?
Yes. tNavigator 26.1 and later include a Local MCP Server as part of AI Assistant. Rock Flow Dynamics says third-party AI applications can use it to query the documentation and manage projects through API functions.
Can Claude connect to tNavigator?
Rock Flow Dynamics says third-party AI applications can connect through the Local MCP Server, and its November webinar promises to show how to connect your own AI client. The public pages don't name specific clients or give configuration.
Does tNavigator's AI Assistant send data outside my company?
By default, Rock Flow Dynamics says requests are processed on a local server inside your own network. It can also connect to external LLM servers through the OpenAI API protocol if you configure that.
Which tNavigator version do I need?
26.1 or later for the Local MCP Server. Version 26.3 brought further AI Assistant upgrades, including a new default language model.
AI-generated models, drawings and results must be checked by a qualified engineer before use.
Sources (checked 4 Nov 2026): tNavigator AI Assistant (Rock Flow Dynamics) · Machine learning and AI with tNavigator (Rock Flow Dynamics) · tNavigator 26.3 key features (Rock Flow Dynamics) · tNavigator product page
tNavigator is a trademark of Rock Flow Dynamics. Tresca is independent and not affiliated with or endorsed by Rock Flow Dynamics.