SPM — Simpler Process Manager

A minimal process manager implementing a lean subset of pm2 features, with some extensions, designed for better development script setup. Lightweight, zero daemon, and PM2-inspired ecosystem config support.

Features

• Start / Stop / Restart — Manage multiple services from a single config
• Multi-instance — Run multiple instances with port/env auto-increment
• Log management — Tail, filter, flush, and rotate logs
• JSON output — jlist for machine-readable status (CI/scripting)
• Log rotation — Size-based rotation and retention with optional background watcher

Installation

npm install -g @hackmd/spm

Or with Bun:

bun add -g @hackmd/spm

Quick Start

1. Create an ecosystem config file (e.g. ecosystem.config.js) in your project:

export default {
  apps: [
    {
      name: 'api',
      script: 'node',
      args: 'server.js',
      instances: 2,
      env: { PORT: '3000' },
      increment_vars: ['PORT'],
    },
    {
      name: 'worker',
      script: 'node',
      args: 'worker.js',
    },
  ],
}

2. Run SPM:

spm start          # Start all services
spm start api      # Start specific service
spm list           # List services and PIDs
spm stop           # Stop all
spm restart api    # Restart specific service
spm logs api -t    # Tail logs

Configuration

| Option | Description | |--------|-------------| | name | Service identifier | | script | Command to run (e.g. node, bun, python) | | args | Arguments passed to the script | | instances | Number of instances (default: 1) | | env | Environment variables | | increment_vars | Env vars to increment per instance (e.g. PORT) | | increment_var | Legacy alias for increment vars (comma-separated string, e.g. PORT,WS_PORT) |

Ecosystem File Field Support

SPM supports PM2-inspired ecosystem files and deeper app objects without failing on extra fields.

• SPM actively uses: name, script, args, instances, env, increment_vars, increment_var
• SPM ignores unsupported fields safely so you can keep partially shared configs across tools
• increment_vars is the preferred form; increment_var is kept as a compatibility alias in SPM

Config file is resolved from ./ecosystem.custom.config.js by default. Override with --config:

spm --config ./my-ecosystem.config.js start

PM2-Inspired Scope

SPM focuses on a small, practical subset of PM2-like workflows for development.

• No daemon model in SPM: processes are started detached and tracked with pid files under ~/.spm2/pids/
• No PM2-style metadata store in SPM: runtime state comes from pid files plus OS process checks
• increment_vars is supported directly; increment_var is also accepted and split by comma
• Additional ecosystem fields can exist in app objects, but SPM only interprets its documented fields

Example for split increment vars:

export default {
  apps: [
    {
      name: 'api',
      script: 'node',
      args: 'server.js',
      instances: 2,
      env: { PORT: '3000', WS_PORT: '4000' },
      increment_var: 'PORT,WS_PORT',
    },
  ],
}

Commands

| Command | Description | |---------|-------------| | spm start [service] | Start all or a specific service | | spm stop [service] | Stop processes (alias: kill) | | spm restart [service] | Restart processes | | spm list | List services and running PIDs | | spm jlist | JSON output of service status | | spm logs [service] | View logs (-t tail, -n lines, -f filter) | | spm flush [service] | Clear log files | | spm rotate start | Start log rotation watcher | | spm rotate stop | Stop rotation watcher |

Log Rotation

Logs are stored in ~/.spm2/logs/. Rotation:

• Rotates when a log exceeds 10MB
• Removes logs older than 3 days
• Run spm rotate start for a background watcher

Shell Integration

Completions require jq. App selector also requires fzf.

Bash

Completions — Source the script (requires bash-completion; on macOS: brew install bash-completion@2):

# After npm install -g @hackmd/spm:
source $(npm root -g)/@hackmd/spm/completions/bash/spm.bash
# Or add to ~/.bashrc:
# source /path/to/spm/completions/bash/spm.bash

App selector — Source the script:

source $(npm root -g)/@hackmd/spm/completions/bash/spm_app_selector.sh
# Then:
spm_app_selector start
spm_app_selector --appName=api restart

Zsh

Completions — Add the completion dir to fpath and ensure compinit runs:

# After npm install -g @hackmd/spm:
fpath=($(npm root -g)/@hackmd/spm/completions/zsh $fpath)
compinit
# Or copy to a dir in $fpath:
mkdir -p ~/.zsh/completions
cp $(npm root -g)/@hackmd/spm/completions/zsh/_spm ~/.zsh/completions/
# Add to ~/.zshrc: fpath=(~/.zsh/completions $fpath)

App selector — Source the script:

source $(npm root -g)/@hackmd/spm/completions/zsh/spm_app_selector.zsh
spm_app_selector start

Fish

Completions — Copy to your Fish config:

mkdir -p ~/.config/fish/completions
cp $(npm root -g)/@hackmd/spm/completions/fish/spm.fish ~/.config/fish/completions/

App selector — Copy the function:

mkdir -p ~/.config/fish/functions
cp $(npm root -g)/@hackmd/spm/completions/fish/functions/spm_app_selector.fish ~/.config/fish/functions/

Usage:

spm_app_selector start
spm_app_selector --appName=api restart
spm_app_selector --config=ecosystem.config.js logs

License

MIT