Cross-platform discovery
NetInsightPro finds the AI agents and MCP servers & tools running on every device reporting into your workspace — by inspecting known MCP client configurations and running processes. Windows, Linux and Android.
AI-Agent & MCP SecurityNew capability
See and secure the AI agents & MCP servers running on your fleet.
AI assistants like Claude Desktop and Cursor now run local tools through MCP (Model Context Protocol) — a fast-growing, mostly invisible endpoint attack surface. NetInsightPro discovers every AI agent and MCP server across your fleet, risk-scores each one with evidence, and lets you block a risky one in a click.
Start free: 14-day Pro trial in the app, no card. Start Pro trial: card on file, 14 days free, cancel anytime. Included on Pro; advanced controls on Enterprise.
- Windows · Linux · Android
- Runs alongside your AV / EDR
- Evidence-backed, not a black box
- Audit-logged
New category · shadow AI
What is MCP — and why it is now an endpoint risk
MCP (Model Context Protocol) is the open standard AI assistants use to run local tools and servers on a machine — reading files, calling APIs, running commands on the user's behalf. Assistants like Claude Desktop, Cursor and VS Code use it every day.
That power is also a new attack surface. MCP servers request permissions and hold secrets; an over-permissioned or malicious one is a supply-chain risk that installs in seconds and shows up in no asset register. Most of it is shadow AI — running on endpoints that security teams cannot yet see.
Incumbent tools do not sell this self-serve yet. NetInsightPro does — and it runs alongside your antivirus / EDR, not instead of it.
What ships today
Discover, inventory, risk-score, drill down, block
Every capability below is live and shipping. The one item still expanding — vulnerability scanning — is called out honestly.
One inventory for the fleet
Every asset in one place: its type (agent / MCP server / MCP tool), the device it runs on, its provider, whether it is active or stale, and when it was last seen — deduplicated to the latest known state.
Risk levels with evidence
Each asset is scored Low, Medium, High or Critical from its declared permissions, provider recognition and secret exposure. The evidence behind every score is shown — not a black-box number.
Evidence & drill-down
Open any finding to see its permissions, declared secrets, configuration fingerprint, last-seen and the exact signals behind its risk — finding → asset → evidence → remediation.
One-click block & quarantine
Block a risky agent or MCP server across your fleet in a single click. On-device quarantine for locally-run assets is rolling out.
Console dashboard + on-device HUD
A security dashboard in the web console for the whole fleet, and an on-device HUD on each machine — with audit logging of what was discovered and what changed.
Vulnerability scanning
Included · rolling outScanning for known CVEs, vulnerable dependencies and malicious packages is included and switching on across the fleet. Per-package results are expanding — today's risk scoring is heuristic and evidence-backed, so you get a defensible signal now and deeper scan detail as it lands.
Pro & Enterprise
Start on Pro. Scale to the fleet on Enterprise.
Pro
Discovery & control- Discovery & inventory across your devices
- Risk levels (Low → Critical) with evidence + drill-down
- One-click manual block / quarantine
- Security dashboard + on-device HUD
- Vulnerability scanning — included, rolling out
Free: 14-day Pro trial in the app, no card. Pro trial: card on file, 14 days free, cancel anytime.
Enterprise
Advanced & automated- Everything in Pro, across your whole fleet
- Advanced policies + automatic blocking
- SOC dashboard for AI-agent & MCP risk
- Audit logging, retention & compliance
- SIEM egress & integrations
Fleet-wide automatic blocking, SOC dashboard, audit + compliance and SIEM egress.
Trust & boundaries
A new layer, not another agent to fear
AI-agent & MCP security is a genuinely new category. We add it honestly: metadata and risk signals, complementing the tools you already run, with an audit trail you can hand to an auditor.
What we see — and what we don't
- We inventory agents and MCP servers — their metadata, permissions and risk signals — not the contents of your prompts or files.
- Runs alongside your antivirus / EDR, adding the AI-agent & MCP layer they do not cover — never replacing them.
- Every discovery and block is audit-logged, so you can show what was found and what changed.
See the AI agents & MCP servers on your fleet today
Start free in minutes, or talk to us about fleet-wide automatic blocking, SOC and compliance on Enterprise.
Included on Pro (discovery, risk, manual controls) · advanced policies, auto-block, SOC, audit & integrations on Enterprise. Runs alongside your AV / EDR.