ST Gate · Our MCP Gateway

One governed door for every MCP tool call.

ST Gate is our own MCP gateway, built in-house — a critical component of Sentinel Telemetry. Every Model Context Protocol client your people run routes through one governed entry point, so an ungoverned tool never reaches your code or your systems unseen.

  • Approved-server registry
  • Full tool-call audit
  • Credential brokering
  • Fail-open
ST Gate routes every MCP client through one governed door Claude Code, Cursor, and VS Code route through ST Gate; approved servers are reachable, unapproved servers are blocked. Claude Code Cursor VS Code ST Gate registry · audit Approved servers Blocked

What ST Gate does.

ST Gate proxies every MCP client through one place you control — and turns an ungoverned sprawl into a single, audited, policy-enforced door.

  • Claude Code
  • Cursor
  • Windsurf
  • VS Code
  • Claude desktop
  • ChatGPT desktop
ST Gate MCP gateway flowchart Multiple MCP clients route through ST Gate, which checks an approved-server registry, audits every tool call, and brokers credentials. Approved servers are reached; unapproved servers are blocked. Claude Code (CLI) Desktop apps Cursor / Windsurf VS Code ST Gate approved-server registry tool-call audit credential broker ◆ one governed door Approved MCP servers reachable + audited Unapproved servers blocked at the gate

Why MCP needs a gate.

The Model Context Protocol lets AI agents and apps call external tools and servers — read files, hit APIs, run actions. That power is exactly why it needs governance. Left ungoverned, every engineer wires up their own MCP servers: unknown tools, reaching into your systems, with no registry and no audit trail.

Shadow tools

Anyone can point an agent at any MCP server. Security has no inventory of what's reachable, and no way to say no.

No audit trail

Tool calls execute directly from the client. When something goes wrong, there's no record of what was called or with what arguments.

Credentials in configs

Tokens end up pasted into local client configs — copied, synced, and leaked far beyond where they should live.

Built in, not bolted on.

Approved-server registry

You decide which MCP servers your people can reach. Anything not on the list is blocked at the gate — no shadow tools, no exceptions you didn't sign off on.

Full tool-call audit

Every tool call — which server, which tool, which arguments, from whom — is recorded and stitched to one employee identity, ready for your SIEM.

Credential brokering

Secrets stay in ST Gate, not in local client configs. The gateway injects credentials at call time, so tokens never sprawl across machines.

Every client, one door

CLIs, desktop apps, Cursor, Windsurf, and VS Code all route through the same governed entry point — one policy, everywhere.

Fail-open by design

If ST Gate is down, slow, or wrong, it never hangs an engineer's agent. Governance in the path, never in the way — the same hard release gate as the rest of the platform.

On your infrastructure

ST Gate runs where ST Hub runs — on-prem or your own cloud, air-gap capable. No tool call is proxied through anyone else's servers.

ST Gate is our own gateway, developed in-house and source-available to licensed customers — your security team can read exactly how every tool call is governed before it runs.

Govern every MCP tool call on infrastructure you own.