The agent
Pure Windows PowerShell 5.1 — already on every supported machine. Runs as a SYSTEM scheduled task with a watchdog that restarts it within 5 minutes.
Shield360 Endpoint monitors, hardens and contains your Windows Servers and PCs from a single self‑hosted console. A lightweight PowerShell agent, Microsoft Defender as the engine, and a prevention layer that heals itself every 15 minutes.
irm https://shield360.biz/install.ps1 | iex
Preview of the Shield360 console with sample data: 248 endpoints, CPU and memory charts, top hosts and open alerts.
Agent check‑in interval. Live CPU, memory, disk and AV status for every machine.
Policy re‑applied automatically. Local tampering is undone before anyone notices.
Attack surface reduction rules, each set to Block, Audit or Off per policy.
Server mirror refresh of Microsoft's signed definition package for offline networks.
An agent on each endpoint, one Node.js server with SQLite, and a browser. No cloud tenant, no per‑seat licence server, no third‑party agent to babysit.
Pure Windows PowerShell 5.1 — already on every supported machine. Runs as a SYSTEM scheduled task with a watchdog that restarts it within 5 minutes.
Node.js 22.13+ with built‑in SQLite. No npm packages to audit or patch. Runs on Windows Server, Windows or Linux.
A fast, focused web UI served with strict security headers. Overview, endpoints, alerts, policies, definitions and audit — nothing you don't need.
Hardware, security posture, inventory and live performance — reported every minute, searchable in seconds.
Shield360 manages Microsoft Defender — built into every supported Windows version — and adds its own prevention layer on top. Configure it once under Protection policies, assign it per endpoint, and pilot safely in Report only mode.
Real‑time, behaviour, script (AMSI), download and cloud protection. Cloud block level, PUA blocking, signature update interval, a daily quick scan and managed exclusions — all from one place.
17 ASR rules — LSASS credential theft, Office macro abuse, obfuscated scripts, PsExec/WMI lateral movement, webshells, USB malware.
The advanced ransomware protection rule plus controlled folder access guard the files that matter.
Defender network protection, plus blocklisted domains via hosts file and IPs via the firewall.
Watches failed logons — RDP, SMB and more — and blocks the attacking IP in the firewall for a set time. Trusted IPs stay on an allowlist.
SHA‑256 hashes are killed and quarantined. IPs, CIDRs and domains are blocked. Paste whole threat feeds.
Firewall on, inbound blocked by default. SMBv1, LLMNR and WDigest off. NLA for RDP, PowerShell logging, optional LSA protection.
Event‑log watch for log clearing, new accounts, privileged group changes, suspicious services, Defender being disabled and attack‑like PowerShell.
Every endpoint on Microsoft's latest definitions — with a signed server mirror for networks that can't reach Microsoft.
Isolate or release, kill a process, quarantine a file, scan a path, remove threats, unblock an IP. And when an endpoint is compromised, it cuts itself off — without waiting for the server.
See isolation in action →Every 15 minutes the agent re‑applies its policy and reports compliance for each individual setting. Someone switched off PUA protection locally? It's back on the next cycle. Settings managed by Group Policy or Intune take precedence and are reported honestly as non‑compliant.
The moment an enabled trigger fires, the agent isolates the endpoint on its own — so it works even while your server is down. All inbound and outbound traffic is blocked, IPv4 and IPv6, internet and LAN. Except one line: back to Shield360, so you stay in control.
Microsoft publishes new Defender definitions several times a day. Shield360 keeps every endpoint current in three layers — including the ones that can't reach the internet.
Runs at the interval set in your protection policy. Default: every 4 hours.
Every 15 minutes. Stale definitions are updated by the agent itself. Above twice the interval it's non‑compliant; above 3 days it raises an alert.
The server fetches Microsoft's full package (mpam-fe.exe, ~200 MB) and keeps it only with a valid Microsoft signature. Isolated PCs and offline networks download it from you — and verify the signature again before installing.
Management tools are high‑value targets. Shield360 is built so a compromised console can't be turned into a remote‑execution platform.
Each agent gets its own random key at enrollment. The server stores only a SHA‑256 hash. Removing an endpoint revokes it instantly.
Remote actions come from a fixed list. The server cannot run arbitrary scripts on your endpoints — by design.
The enrollment token only registers new machines. Rotate it any time without touching enrolled agents.
Admin passwords hashed with scrypt. HttpOnly, SameSite=Strict sessions. Five failed logins lock the source IP for five minutes.
The console ships with CSP, frame denial and nosniff. HTTPS with your own internal CA certificate is a single command.
Config and key live in C:\ProgramData\Shield360, readable only by SYSTEM and Administrators. Full audit log of every action.
Sensible rules out of the box, thresholds you can tune in config.json. When the condition clears, the alert closes on its own — no stale queue to triage on Monday.
No active antivirus · Defender real‑time off · Threat detected · Auto‑isolated
Endpoint offline · Firewall profile disabled · Failed‑logon spike · Pending security updates
Disk ≥ 90% · Sustained CPU ≥ 90% · Memory ≥ 90% · Outdated signatures · Workstation without BitLocker
Restart pending
One command installs the server — Node.js check, verified download, HTTPS certificate and auto‑start included. Then the console hands you ready‑to‑paste agent commands with your server URL and token already filled in.
Paste the one‑line command from the Deploy page into an elevated prompt.
Within a minute the machine appears under Endpoints. Assign a policy, start in Report only, then switch to Enforce.
Everything the installer uses, for air‑gapped networks or your own review.
Any Windows Server or Windows machine (Linux works too) with Node.js 22.13 or later. No npm packages.
Windows Server 2016 / 2019 / 2022 / 2025 or Windows 10 / 11 with built‑in Windows PowerShell 5.1.
GPO startup script out of the box. Intune, SCCM or any RMM tool can run the same command.
Still curious? The README and testing guide ship with every copy, or write to [email protected].
No. Shield360 uses Microsoft Defender — already built into every supported Windows version — as its engine, and manages it centrally. It also reports third‑party AV where present.
Not by default. The "Servers: alert only" setting is on out of the box. Servers raise a critical "Compromise detected" alert and you isolate them with one click when you choose to.
Endpoints keep enforcing their last policy and still isolate themselves when a trigger fires — the decision happens on the agent. When the server returns, everything reconciles on the next check‑in.
It checks for Node.js 22.13+ (installing the LTS with winget if missing), downloads the release package from shield360.biz, verifies its SHA‑256 checksum, extracts it to C:\Shield360 Endpoint, creates an HTTPS certificate and registers the server to start automatically. Upgrades keep your config and data. Read the script before you run it.
Use Report only mode. The agent measures compliance for every setting but changes nothing, and raises alerts instead of isolating. Switch to Enforce when you're happy.
Yes. Settings managed by GPO or Intune take precedence, and Shield360 shows them as non‑compliant so you can see exactly where policies disagree.
No. Remote actions come from a fixed allow‑list — scans, signature updates, process and connection listings, restarts with a two‑minute user warning, isolation and the response actions. Nothing else.
Self‑hosted. Dependency‑free. Ready in minutes.
Questions about licensing or a rollout? Email [email protected]