finds.dev

DragoQC/MeshAll

C# orchestration, Blazor Server and Tailwind UI, SQLite, and maintained native remote-access technology

View on GitHub

CMITupdated Oct 2026

// the take

MeshAll is a from-scratch C#/ASP.NET Core management plane (Blazor Server UI, SQLite storage) bolted onto the retained native MeshCentral agent, agent-core JS modules, and browser viewers (xterm.js, noVNC, Guacamole/FreeRDP for RDP/SSH). It's aimed at people running MeshCentral for remote device management who want to drop the Node.js orchestrator but keep the native remote-access stack it's built on.

Keeping the native C/C++ agent and the existing browser viewers instead of rewriting them in C# is the right call — that code is mature and cross-platform, and reimplementing a KVM/remote-control agent from scratch would just introduce new bugs for no benefit. The per-session screen-share consent model (explicit ask-before-show, no remembered 5-minute approval in strict mode, deny on timeout) is a real improvement over MeshCentral's unattended-by-default behavior. SQLite keeps the ops footprint small for a self-hosted single-server tool. The README is upfront that orchestrator/UI parity with MeshCentral isn't proven yet and points at a COMPATIBILITY.md tracking matrix instead of just claiming feature parity.

Zero stars, zero forks, no track record — for software that does native-agent certificate authentication and grants full remote control of machines, that means no outside security review and an install path that's curl-pipe-to-bash as root. The native agent binaries for all four platforms are committed to the repo as prebuilt binaries rather than built reproducibly from the vendored agent-native source in CI, which is exactly the kind of supply-chain risk you don't want in a remote-access tool regardless of the included provenance.json. The project's own docs admit the C# orchestrator hasn't been verified against the original MeshCentral server's behavior, so this is pre-production for a security-sensitive use case. It's also now a fork that has to track two moving targets — upstream MeshCentral's native agent/viewer fixes and its own new C# server — with apparently one maintainer responsible for porting security fixes across both.

We read the README, the directory tree and a key source file in October 2026. Nobody pays to be listed. How we judge →

// at a glance

Stars
0
Forks
0
Language
C
License
MIT
Last commit
23 Oct 2026

// nearby

If MeshAll isn't quite it.

All repos →