SkillOfficial (tool developers)
reviewing-openmc-code skillby OpenMC developers
The OpenMC project's own code-review skill: checks a branch or pull request against OpenMC's contribution criteria.
1.1k· updated 1 day ago
OpenMC MCP servers and AI agent skills
OpenMC has two community MCP servers and two skills. PromptMC validates OpenMC XML against typed schemas before it runs, plots and parses simulations, and openmc-validator-mcp runs pre-flight checks on materials, geometry and tallies. A community skill makes Claude Code read your local OpenMC docs before writing code, and the OpenMC project ships its own code-review skill for contributors. Running simulations through any of them needs OpenMC and your own cross-section data.
- Includes Open source's official server.
- 2 MCP servers and 2 agent skills listed.
Open source2 servers · 2 skillsChecked 4 Nov 2026
Integrations for OpenMC
Official
Community and commercial servers
MCP serverCommunity
openmc-validator-mcpby zxkjack123
Pre-flight checks for OpenMC models over MCP: 14 tools that catch material, geometry, tally and data errors before a run.
LocalRead-only0· updated 3 months ago
MCP serverCommunity
PromptMCby rjonace
Puts a schema-checking gate between an AI assistant and OpenMC, then runs, plots and parses validated decks over MCP.
LocalmacOS, LinuxCan modify models2· updated 5 months ago
Agent skills
SkillCommunity
Claude OpenMC Skills (openmc-api)by cn-skywalker
An openmc-api skill that makes Claude Code read your local OpenMC docs before writing model code, with model templates.
0· updated 8 months ago
Listings are sorted by origin, then by recent activity. GitHub data is refreshed nightly. Nothing here is ranked.
How AI connects to OpenMC
An MCP server or skill can only do what OpenMC lets programs do. OpenMC exposes 2 automation interfaces:
- OpenMC Python API (Python): Most OpenMC work happens in Python: you build materials, surfaces, cells, lattices, settings and tallies as objects, export them to XML, run the transport solver, and read statepoint files back for k-effective and tally results. The same package handles depletion and multigroup cross sections. Both MCP servers here sit on this layer, and so does the community skill's set of templates. Documentation
- C API (openmc.lib) (Python): The OpenMC repository's agent instructions describe a C++17 transport core with C API bindings wrapped through ctypes in openmc.lib, which give runtime control of a simulation from Python. That is the route for coupling OpenMC to other codes, for example in multiphysics loops, rather than for ordinary model building. Documentation
What Open source offers
The OpenMC developers have added AI-agent support to the repository itself: an AGENTS.md file describing the codebase for coding agents, and a reviewing-openmc-code skill (first added in March 2026) that checks contributions against OpenMC's review criteria. These target people developing OpenMC. The project hasn't released an MCP server; PromptMC and openmc-validator-mcp are independent.
Checked 4 Nov 2026. Sources: reviewing-openmc-code SKILL.md (openmc-dev/openmc), AGENTS.md (openmc-dev/openmc)
What you can do with OpenMC and AI
- Draft a pin cell or fuel assembly model from a description, validate the XML against schemas, and fix reported errors before any run starts.
- Check that every nuclide and temperature in a model exists in your cross_sections.xml and that tally filters match real cells and mesh extents.
- Run a simulation, make 2D geometry slice plots and report k-effective and tally means with their relative errors in the chat.
- Convert an OpenMC neutron spectrum tally into a FISPACT-II spectrum file for activation work.
- Review a branch or pull request against OpenMC's own contribution criteria if you develop OpenMC itself.
Taken from the listings' own documentation. Check each listing for exactly what it supports.
Installing
The listings here document setup for Claude Code, Claude Desktop, Cursor, Google Antigravity and VS Code (GitHub Copilot). Install steps, copied from each project's own documentation, are on each listing's page. Most OpenMC integrations run on your own computer next to OpenMC, so you need OpenMC installed and licensed on the same machine unless the listing says otherwise.
Related tools
Questions
Is there an OpenMC MCP server?
Yes, two community ones. PromptMC exposes validation, templating, running, plotting and statepoint parsing; openmc-validator-mcp focuses on pre-flight checks and post-processing. Neither comes from the OpenMC developers.
Is there an official OpenMC agent skill?
Yes, for contributors. The OpenMC repository includes a reviewing-openmc-code skill and an AGENTS.md file. For building models, use the community openmc-api skill or one of the servers.
Do I need nuclear data to use these tools?
For schema validation, no. To run simulations, plot geometry or check cross-section availability you need OpenMC installed and OPENMC_CROSS_SECTIONS pointing at a cross_sections.xml.
Can an AI agent design a reactor with OpenMC?
These tools help an agent build and check input faster, but PromptMC's author states the goal is safer, faster iteration, not autonomous reactor design. Review every model and result.
AI-generated models, drawings and results must be checked by a qualified engineer before use.
Sources (checked 4 Nov 2026): OpenMC documentation · OpenMC Python API reference · openmc-dev/openmc repository · OpenMC product page
OpenMC is developed by its own project (Open source). Tresca is independent and not affiliated with or endorsed by the OpenMC project.