INSIDE MESHPLOY

WireGuard connects your machines.
K3s runs your workloads.

Meshploy brings routing, builds, deployments, and access control together—with mesh-only nodes for services running outside Kubernetes.

Explore the layers ↓
01 / PLATFORM COMPONENTS

Built on clear foundations.

Select a layer to stop cycling. Press Play to resume.
Functional groups, not sequential steps.

Interfaces Console · CLI · AI agent / MCP Control plane API + permissionsPlatform database Build & runtime Build → registryK3s workloads Networking Gateway + routingPrivate mesh Integrations Git · Object storageNotifications
Interfaces highlighted · Neighboring groups shown for context
Interfaces

Choose how you work.

The console, CLI, and AI agent tools connect to the same platform API. Permissions scope what each identity can do.

  • React console
  • Go CLI
  • MCP tools
Interfaces Console · CLI · AI agent / MCP Control plane API + permissionsPlatform database Build & runtime Build → registryK3s workloads Networking Gateway + routingPrivate mesh Integrations Git · Object storageNotifications
Control plane highlighted · Neighboring groups shown for context
Control plane

Coordinate the platform.

The API checks access and coordinates resource operations. Platform configuration and state are stored in PostgreSQL.

  • API
  • Auth & RBAC
  • Platform PostgreSQL
Interfaces Console · CLI · AI agent / MCP Control plane API + permissionsPlatform database Build & runtime Build → registryK3s workloads Networking Gateway + routingPrivate mesh Integrations Git · Object storageNotifications
Build & runtime highlighted · Neighboring groups shown for context
Build & runtime

From source to running workloads.

Build jobs produce container images. Workers pull them from a registry, and K3s schedules workloads according to capacity and placement rules.

  • Build jobs
  • Image registry
  • K3s / Kubernetes
Interfaces Console · CLI · AI agent / MCP Control plane API + permissionsPlatform database Build & runtime Build → registryK3s workloads Networking Gateway + routingPrivate mesh Integrations Git · Object storageNotifications
Networking highlighted · Neighboring groups shown for context
Networking

Reach the right service.

The gateway routes public traffic to selected endpoints over the private mesh. Existing services can participate without becoming Kubernetes workloads.

  • Caddy + proxy
  • WireGuard / Headscale
  • CoreDNS
Interfaces Console · CLI · AI agent / MCP Control plane API + permissionsPlatform database Build & runtime Build → registryK3s workloads Networking Gateway + routingPrivate mesh Integrations Git · Object storageNotifications
Integrations highlighted · Neighboring groups shown for context
Integrations

Connect the services around it.

Git providers supply source and deployment triggers. Backup jobs write to configured object storage, while notifications deliver platform events.

  • Git providers
  • S3-compatible storage
  • Notifications
Read the documentation ↗
View full architecture

From an action to a running service.

Management requests, builds, public traffic, and external integrations follow different paths.

CONTROL PLANE RUNTIME + PRIVATE WIREGUARD MESH EXTERNAL SERVICES ConsoleCLIAI agent / MCPAPI + access controlPlatform PostgreSQLPlatform servicesGit providersBackup storage (S3)Notification channelsBuild jobsImage registryK3s workloadsPublic requestsCaddy + Meshploy proxyExisting services pushpull HTTPS / TCP Headscale coordinates membership · Nodes communicate over WireGuard

Simplified system view. Lines show calls and data flow, not physical machine boundaries. DNS, metrics, and detailed job internals are omitted for readability.

02 / FROM SOURCE TO SERVICE

A deployment’s journey.

Build on one machine. Run on another. The private mesh connects the work.

  1. 01

    Source

    Git repository

  2. 02

    Build

    Source → container image

  3. 03

    Registry

    Store and distribute

  4. 04

    Worker

    Roll out the workload

  5. 05

    Route

    Publish an endpoint

Pre-built images skip the build step. Build nodes must be online. Replica placement depends on capacity and scheduling rules; increasing replicas is not automatic scaling.

03 / CONNECTED, WITH DIFFERENT ROLES

Every node has a purpose.

The gateway provides the public entry point. Other machines contribute the capabilities you assign to them.

Gateway

Public routing and the platform’s entry point. Direct selected traffic to workloads or existing endpoints.

Worker

Runs Kubernetes workloads. Applications use the resources available on eligible nodes.

Build node

Produces container images from your source. A supported local machine can participate when it is online.

Mesh-only node

Connects existing services over the private network without becoming a Kubernetes worker.

04 / SHARED RESPONSIBILITIES

Know what you control.

Meshploy coordinates

Deployment workflows, resource configuration, mesh membership, routes, access permissions, and configured backup jobs.

You operate

Server provisioning, host security, firewall rules, capacity, DNS prerequisites, and backup storage. Verify your recovery procedures and keep the underlying machines available.

See the system at work.

Explore a sample workspace, or set up Meshploy on your infrastructure.