Would this bed mesh → bed screw adjustment tool be useful for Creality printers & users?

Flagship K Series
,
1

Hi all,

I’ve made a small open-source bed levelling assistant and wanted to see whether it would be useful to other Creality users.

Repo:

I originally made this because I was wary of modifying Klipper on my K2 Pro just to add a screw tilt style workflow. I wanted an external tool that could use the bed mesh data I already had and give practical bed screw adjustment guidance.

It started fairly simple, but it has gradually grown arms and legs.

At the moment, the tool takes a regular bed mesh and tries to separate the overall bed tilt from local bed warp. It then gives per-screw adjustment suggestions based on the fitted bed plane.

I’ve also started looking at chamber and bed thermal response, because I wanted to understand whether the mesh changes I was seeing were purely mechanical, or partly caused by the printer changing as the bed and chamber heated up.

I’m mainly looking for feedback from people using Creality printers with manual bed screws and mesh probing.

Questions:

  • Would this be useful in your bed-levelling workflow?
  • What printer and firmware are you using?
  • Do you currently use any screw tilt or screw adjustment workflow?
  • Would you prefer an external tool rather than modifying Klipper/printer config?
  • Can you easily copy/export your bed mesh?
  • Have you noticed your mesh changing between cold, warm, and fully heat-soaked conditions?
  • Would a Windows .exe be enough, or would another format be better?
  • What would make the tool easier to trust?
  • Are the screw adjustment results clear enough to act on?

Known limitations:

  • It needs a rectangular mesh.
  • Correct screw positions, screw pitch, bed size, and mesh bounds matter.
  • Thermal behaviour is complicated and I’m still treating that side cautiously.
  • The physical-response model is advisory rather than the main recommendation method.
  • The Windows executable is currently unsigned, so SmartScreen may warn.

I’m not claiming this is finished or perfect. I’m trying to find out whether it solves a real problem before I spend more time polishing it.

Any feedback, criticism, or real mesh data would be appreciated!