Skip to content

Command Palette

Search for a command to run...

Version 1.1.0

Phelix documentation

Build, run, deploy, and monitor Go and Rust applications across one machine or an entire server fleet.

What is Phelix?

Phelix is a Go-based CLI for managing the complete lifecycle of Go and Rust applications. It combines local process management with versioned releases, encrypted configuration, safe deployment strategies, container builds, and multi-server observability. It targets Linux, with experimental Windows support.

As of v1.1.0, Phelix is also a controlled execution layer for coding agents: a built-in MCP adapter, a fail-closed execution-authorization boundary, durable agent sessions, and a versioned --json machine contract on every core command. An agent proposes what to deploy; Phelix decides whether this caller may execute that exact plan, and a human decides who may sign it off.

Quick start

Terminal
# Initialize project config (creates phelix.yaml)phelix init # Build and run an app from the current directory (auto-detects Go or Rust)phelix build myapp --port 8080 # Rebuild after a code change (classic by default)phelix rebuild myapp # One-time migration from classic to the proxy (brief downtime)phelix stop myappphelix rebuild myapp --blue-green # Ship a change to 5% of traffic first, verify, then promotephelix rebuild myapp --canary 5 # Diagnose project compatibility (PORT usage, config, toolchain)phelix doctor # List all managed appsphelix list # Roll back to the previous versioned buildphelix rollback myapp

Core workflow

  1. Build
    Auto-detect Go or Rust, compile the project, create a numbered version, record a Build Report, and start the app.
  2. Deploy
    Use classic replacement, blue-green switching, rolling replicas, or verified canary and progressive rollouts.
  3. Observe
    Inspect status, logs, health checks, server metrics, and proxy routing.
  4. Recover
    Roll back to a retained version and restore its paired encrypted environment snapshot.

A typical lifecycle: build then optional zero-downtime rebuild, with rollback as the safety net and stop / remove to retire an app. Every build is versioned (v1, v2, ...) and can be tagged.

Capabilities

AreaIncluded in v1.1.0
RuntimeGo and Rust auto-detection
DeploymentsClassic, blue-green, rolling, canary, and progressive
RecoveryVersioned binaries, tags, dry-run previews, and audited rollback
SecurityAES-256-GCM environment and registry credentials
Build intelligenceAutomatic build reports and regression alerts, fully offline
DistributionDocker, Compose, and matrix builds with release manifests
Docker runtimeRun app instances as containers under the zero-downtime proxy
OperationsHealth checks, logs, proxy, and opt-in multi-server monitoring
Resource controlPer-instance CPU and memory limits via cgroups v2 (Linux)
Git webhooksHMAC-authenticated pushes deploy exact commits with durable job history
Automation & agentsMCP adapter, agent execution sessions, and two-step plan-based execution
AuthorizationOptional fail-closed execution gate with plan-bound, human-signed approvals
Machine interfaceA versioned --json envelope on every core command, for scripts and agents