Skip to content
Tresca

Serpent AI integrations

No MCP server or agent skill for Serpent is listed yet (checked 4 Nov 2026). No MCP server or agent skill targets Serpent as of November 2026. The nearest tools are INL's Virtual Test Bed Assistant, which can find and explain VTB models built with Serpent but can't author Serpent input, and serpentTools, a community Python package for reading Serpent output. Serpent is developed by VTT in Finland; RSICC's catalogue lists Serpent 2.2.5 for licensed distribution.

VTTNone listed yetChecked 4 Nov 2026

How AI connects to Serpent

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

  • serpentTools (community output parsers) (Python): serpentTools, maintained by the CORE group at Georgia Tech under the MIT licence, parses Serpent output files into Python objects and installs with pip or conda. An agent working with Serpent could use it to read results instead of writing its own parsers. Its README points to VTT's Serpent site for distribution and licensing details. Documentation
  • Serpent input files (Input files): RSICC's catalogue describes Serpent 2 as a continuous-energy Monte Carlo reactor physics and burnup code. Models are set up in input files rather than through an API, so a files-only agent on a licensed machine is the likely form of any integration. VTT's official documentation could not be reached from our checker on 4 November 2026, so no link is given here yet.

What VTT offers

No AI assistant or MCP server from VTT for Serpent was found in this check, and the Serpent site itself could not be reached from our checker. The only agent tool that mentions Serpent is INL's VTB Assistant, which can locate and explain Virtual Test Bed models whose codes include Serpent but supports input generation only for ten MOOSE-based applications.

Checked 4 Nov 2026. Sources: vtb-docs SKILL.md (idaholab/virtual_test_bed), CORE-GATECH-GROUP/serpent-tools README

What an integration for Serpent could do

  • Write a pin cell or assembly input from a description and check material and geometry definitions before a burnup run.
  • Read depletion and detector output with serpentTools and tabulate k-infinity, nuclide densities and detector responses by burnup step.
  • Compare a Serpent result against an OpenMC or MCNP model of the same problem and list where they differ beyond statistical uncertainty.
  • Find Virtual Test Bed models that used Serpent to generate cross sections for a reactor concept you are studying.

Know of one, or building one?

If an MCP server or agent skill for Serpent 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

OpenMCOpen source2 servers · 2 skillsMCNPLANL1 skillSCALEORNLNone listed yetMOOSEINL1 server

Questions

Is there a Serpent MCP server?

No. Nothing was found on 4 November 2026. OpenMC, a different Monte Carlo code, has two community MCP servers if you can switch codes for a study.

How is Serpent distributed?

Serpent is developed by VTT, and RSICC's software catalogue lists SERPENT2.2.5. Check VTT's Serpent site or RSICC for the licence terms that apply to you.

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

Sources (checked 4 Nov 2026): CORE-GATECH-GROUP/serpent-tools README · serpentTools documentation · RSICC software catalogue, S

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