Claude Can Now Reach Your Internal Database: MCP Tunnels Launch

EDIBLOG · AI · NEWS · 2026.05.21

Claude Can Now Reach Your Internal DB — MCP Tunnels Launch

On May 19, at Code with Claude London, Anthropic unveiled two new security features for Managed Agents. There’s now a way to wire agents into internal systems without poking a hole in your firewall.

MCP Tunnels Self-Hosted Sandboxes research preview 2026.05.19 Code with Claude

01 · HEADLINEBoth features shipped together

Anthropic added two new security features to Claude Managed Agents: MCP Tunnels and Self-Hosted Sandboxes. They were announced as a pair on the same day, and both push in the same direction — stronger enterprise security and privacy.

Before
Public
Agent → routed over public internet
MCP server must be exposed publicly
Tool execution on Anthropic infra
Firewall needs inbound rules
2026.05.19
shipped
After
Private
Inside your network + your own infra
MCP server stays private
Tool execution on customer infra
Firewall one outbound only

The two ship at different stages, though. Self-Hosted Sandboxes is in public beta; MCP Tunnels is a limited research preview, so it requires requesting access.

02 · MCP TUNNELSHow to call an agent inside the firewall

In one line, MCP Tunnels lets Claude call your internal MCP server without putting it on the public internet. You deploy a lightweight gateway inside your network, and that gateway opens a single encrypted outbound connection toward Anthropic.

▮ THE OLD BURDEN until 2026.05.18

“To use an internal DB, ticketing, or KB as an agent tool, you needed a public endpoint.”

The burden: adding inbound firewall rules, the risk of exposing auth and credentials, and rounds of security-team review. The tighter the security gate — think finance or healthcare — the more often PoCs stalled there.

With Tunnels, you don’t need to add inbound firewall rules. Allowing a single outbound connection is enough. Internal databases, private APIs, knowledge bases, and ticketing systems become agent tools as-is. Per VentureBeat’s reporting, the core value is that credentials never cross the public internet.

03 · SANDBOXESOnly tool execution moves to your infra

Self-Hosted Sandboxes is a split architecture. The agent loop (orchestration, context, error recovery) stays on Anthropic’s infra, while only tool execution moves to customer infra. Sensitive files, packages, and services never leave your network.

▮ LAUNCH PARTNERS (4 AT RELEASE)
  • Cloudflare — microVM + zero-trust networking, outbound traffic controltrust boundary and traffic control angle
  • Daytona — long-running stateful environments, SSH / preview URL accessfor workflows that need to keep state
  • Modal — AI-workload focused, CPU/GPU scale allocationcompute-heavy work like long builds and image gen
  • Vercel — sandbox isolation + VPC peering + credential injectioncredential injection at the network boundary

You control the resource sizing and the runtime image. For compute-heavy tasks like long builds or image generation, you can directly allocate the CPU, memory, and capacity you need.

04 · DATAThe numbers at a glance

NEW
FEATURES
2
Tunnels + Sandboxes
SANDBOX
PARTNERS
4
CF / Daytona / Modal / Vercel
TUNNELS
STATUS
RP
research preview · apply
SANDBOX
STATUS
β
public beta · now

Existing Managed Agents integration code doesn’t need to change. Anthropic says a single config change lets you move between Anthropic infra and customer infra. In other words, there’s no new SDK call or separate migration procedure.

05 · POSITIONWhat changes for teams on the ground

▮ POSITION · OPEN 2026.05.21 KST

The tighter your data-exfiltration gate — think finance or healthcare — the more these two features matter. PoCs that used to die in security review now reach the starting line with one outbound connection and a config change.

▮ GOOD FIT

Workflows where an internal DB, issue tracker, or KB has to become an agent tool. Compliance environments where data must not leave the network.

▮ CAVEAT

Tunnels is a limited research preview, so it requires requesting access. General availability, pricing, and plans aren’t separately spelled out on the official page.

“What kept the agent stuck wasn’t model capability — it was one line of firewall. Turning that line into a single outbound connection is the heart of this announcement.”

— Eddie · 2026.05.21

This isn’t an update to the model — it’s an update to the boundary. I expect more teams to take a fresh look at whether they can bring this in-house.

Disclaimer: A roundup of an external announcement. No ads, no affiliates. Figures and statuses reflect the announcement date and may change later.


This post is part of an English mirror of a Korean dev experiments blog. Cross-posted for the global developer audience.

Leave a Comment