Sielum
Open Source · AI Governance

AI agent monitoring. Finally.

Know every AI agent running on your developer machines — before your next SOC 2 or EU AI Act audit asks you to prove it. Install in 15 minutes. Free for up to 5 endpoints.

Install in 15 minutesBook a Demo
mTLS Device Auth
SOC 2 Aligned
EU AI Act Ready
GDPR Compliant

Monitors

Claude CodeGitHub CopilotCursorChatGPT DesktopWindsurfGemini CLI+ more

How It Works

Up and running in 15 minutes

No demo call required. No sales cycle. No enterprise deployment project.

01

Install the agent

One command. Runs natively on Linux, macOS, and Windows — no Docker, no sidecar, no proxy. The agent needs direct OS access to see what's actually running.

curl -sSL https://get.sielum.io/install.sh | sh
View documentation →
02

Agents register automatically

Each agent enrolls via mTLS in under a minute. Device certificates are issued automatically — no shared secrets, no manual config, no token sprawl.

03

See everything. Instantly.

Every AI tool running on every endpoint. Every API connection. Every MCP server. Live in the dashboard — within 15 minutes of installation.

Platform Capabilities

Everything you need to govern AI agents

Built for security teams who need real answers, not dashboards full of noise.

Endpoint Visibility

See every AI agent running on every device. Processes, configs, and MCP servers in one place. No blind spots.

API Connection Monitoring

Know which AI APIs each agent calls. Block unauthorized endpoints with one click. Real-time connection telemetry.

MCP Server Control

Discover and govern Model Context Protocol servers before they become a shadow IT problem. Approve, block, or audit.

Policy Engine

Define rules. Get alerts. Enforce configuration standards across your entire fleet automatically — at scale.

Compliance Ready

GDPR-compliant data controls, SOC 2-aligned audit logs, and EU AI Act reporting built in. One-click export.

Zero-Trust Agent Auth

Every endpoint agent authenticates with mTLS device certificates. No shared secrets. No token sprawl. Ever.

Why Sielum

Built differently. On purpose.

Most AI monitoring tools require a proxy, a vendor account, or a sales call. Sielum requires none of those.

Proxy-free architecture

No man-in-the-middle. The agent monitors locally on the endpoint — no traffic is rerouted through our servers. You keep full control.

No proxy required

Agents report directly to your server. No traffic rerouting, no man-in-the-middle, no TLS inspection. Your developers' AI traffic stays private.

Self-hosted or cloud — your choice

Run the server on your own infrastructure or use our managed cloud. Switch at any time. Your data never has to leave your network.

Air-gap and on-prem capable

Full functionality without any internet dependency. Built for organisations with strict data residency requirements or disconnected environments.

Self-service — no demo call required

Install the agent, connect to the dashboard, see results. The free tier works without contacting sales, signing a contract, or waiting for onboarding.

Use Cases

Real problems. Real answers.

Security teams monitoring developer endpoints use Sielum to solve the problems that keep CISOs up at night.

01

Shadow AI Detection

Your developers are using AI tools you don't know about. We surface every instance — automatically.

AI tools spread faster than IT policy. Claude Code, Copilot, Cursor, ChatGPT Desktop — each one a potential data-exfiltration vector. Our agent discovers them on every endpoint, reports their MCP connections, and alerts you the moment a new tool appears.

Endpoint ScanningProcess DetectionInstant Alerts
02

Pre-Audit Compliance

SOC 2 or EU AI Act audit coming up? Export a full AI agent inventory with one click.

Auditors want evidence. Our platform generates structured reports in CSV or PDF format: which AI tools run on which endpoints, which APIs they call, which MCP servers are active, and your enforcement history. All timestamped, all auditable.

SOC 2 AlignedEU AI ActOne-Click Export
03

MCP Security

MCP servers can access filesystems, databases, and APIs. Know what's connected before it's a breach.

Model Context Protocol extends AI agents with powerful tool access. A single misconfigured MCP server can expose your entire codebase or database to an AI model. We inventory every MCP server, classify its tool access, and let you approve or block them by policy.

MCP DiscoveryTool ClassificationPolicy Enforcement

What teams are saying

Built for teams who need real answers

The ability to enforce which LLM providers our engineers can reach — fleet-wide, from a single policy — is exactly what we were missing.

Director of Infrastructure

We had no visibility into what our AI agents were actually doing. Sielum was the first tool that gave us a clear answer to that question — and made our compliance requirements manageable.

Director of Internal IT

Our developers had highly individual AI setups — a different mix of tools on every machine. Sielum gave us a single view across all of it and let us enforce consistent controls for the first time.

Software Engineering Manager

FAQ

Common questions

Does the agent read my code or AI prompts?

No. The agent monitors processes and network connections only — it records which AI tools are running and which APIs they connect to. It never reads file contents, prompt text, or AI responses.

Can I self-host everything?

Yes. The Free tier is fully self-hosted — server, dashboard, and database run on your own infrastructure. Your data never leaves your network. Managed cloud tiers are available if you prefer not to operate the server yourself.

What counts as an endpoint?

One endpoint is one developer machine (laptop or workstation) running the Sielum agent. The Free tier covers up to 5 endpoints at no cost.

See all frequently asked questions

From the Blog

Security insights

All posts

Introducing Sielum

Shadow AI is the new Shadow IT. Here's how we built a platform to give enterprise security teams full visibility over every AI agent running in their organization.

Read more

MCP Security: What Every CISO Needs to Know

Model Context Protocol gives AI agents access to filesystems, APIs, and databases. Most security teams have no visibility into what's connected. Here's how to fix that.

Read more

Start free. Scale when you need it.

Deploy the agent on your dev machines today — no sales call required.

Free
€0
Up to 5 endpoints
Starter
€500/mo
Up to 50 endpoints
Growth
€1.000/mo
Up to 200 endpoints
Business
Custom
Unlimited endpoints
Install FreeSee all plansTalk to Sales

Air-gap / on-prem deployment available for Enterprise. Contact us