magpie-mags 1.3.8

Mags SDK - Execute commands and scripts on Mags Sandboxes

Mags Python SDK

Execute scripts on Magpie's instant VM infrastructure from Python.

Install

pip install magpie-mags

Quick Start

from mags import Mags

m = Mags(api_token="your-token")

# Run a script and wait for the result
result = m.run_and_wait("echo 'Hello from a VM!'")
print(result["status"])    # "completed"
print(result["exit_code"]) # 0
for log in result["logs"]:
    print(log["message"])

Authentication

Pass api_token directly, or set one of these environment variables:

export MAGS_API_TOKEN="your-token"
# or
export MAGS_TOKEN="your-token"

Usage

Run a Script

# Fire-and-forget
job = m.run("apt install -y ffmpeg && ffmpeg -version")
print(job["request_id"])

# Run and wait for completion
result = m.run_and_wait(
    "python3 -c 'print(sum(range(100)))'",
    timeout=30,
)

Persistent Workspaces

# First run: creates the workspace
m.run_and_wait(
    "pip install pandas && echo 'setup done'",
    workspace_id="my-project",
    persistent=True,
)

# Second run: reuses the workspace (pandas is already installed)
m.run_and_wait(
    "python3 -c 'import pandas; print(pandas.__version__)'",
    workspace_id="my-project",
)

Always-On VMs

# VM that never auto-sleeps — stays running 24/7
job = m.run(
    "python3 server.py",
    workspace_id="my-api",
    persistent=True,
    no_sleep=True,
)
# Auto-recovers if the host goes down

Enable URL / SSH Access

job = m.run("python3 -m http.server 8080", persistent=True)

# HTTP access
access = m.enable_access(job["request_id"], port=8080)
print(access["url"])

# SSH access
ssh = m.enable_access(job["request_id"], port=22)
print(f"ssh root@{ssh['ssh_host']} -p {ssh['ssh_port']}")

Upload Files

file_ids = m.upload_files(["data.csv", "config.json"])
result = m.run_and_wait(
    "ls /uploads && wc -l /uploads/data.csv",
    file_ids=file_ids,
)

Cron Jobs

cron = m.cron_create(
    name="nightly-backup",
    cron_expression="0 0 * * *",
    script="tar czf /workspace/backup.tar.gz /data",
    workspace_id="backups",
)

jobs = m.cron_list()
m.cron_update(cron["id"], enabled=False)
m.cron_delete(cron["id"])

Check Usage

usage = m.usage(window_days=7)
print(f"Jobs: {usage['total_jobs']}, VM seconds: {usage['vm_seconds']:.0f}")

API Reference

Method	Description
run(script, **opts)	Submit a job (persistent, no_sleep, workspace_id, ...)
run_and_wait(script, **opts)	Submit and block until done
new(name, **opts)	Create a VM sandbox (persistent=True for S3)
find_job(name_or_id)	Find a running/sleeping job by name or workspace
exec(name_or_id, command)	Run a command on an existing VM via SSH
stop(name_or_id)	Stop a running job
url(name_or_id, port)	Enable public URL and return full URL
resize(workspace, disk_gb)	Resize a workspace's disk
status(request_id)	Get job status
logs(request_id)	Get job logs
list_jobs(page, page_size)	List recent jobs
update_job(request_id, **opts)	Update job config (startup_command, no_sleep)
enable_access(request_id, port)	Enable URL or SSH access
usage(window_days)	Get usage summary
upload_file(path)	Upload a file, returns file ID
upload_files(paths)	Upload multiple files
url_alias_create(sub, ws_id)	Create a stable URL alias
url_alias_list()	List URL aliases
url_alias_delete(sub)	Delete a URL alias
cron_create(**opts)	Create a cron job
cron_list()	List cron jobs
cron_get(id)	Get a cron job
cron_update(id, **updates)	Update a cron job
cron_delete(id)	Delete a cron job

Links

• Website: mags.run
• Node.js SDK: npm install @magpiecloud/mags
• CLI: npm install -g @magpiecloud/mags