LC
Liam ChungApril 20, 2026 ยท 4 min read

MCP Security: What Actually Matters in Production

MCP Security: What Actually Matters in Production

MCP security is not mostly a debate about whether the protocol is good. It is a question of what trust boundary you create every time a model can call a server, read a resource, or trigger a tool. Once you frame it that way, the real controls become much clearer.

Why teams misread the risk

Many teams treat MCP as a directory of useful servers. That misses the harder part. MCP is a contract layer between clients and capability providers. If a model can invoke those capabilities inside a workflow that mixes untrusted content and real side effects, trust design becomes critical.

That is why the official materials spend so much time on authorization, tool confirmation, and attack surfaces like confused deputy problems.

The controls that matter first

Capability minimization matters more than ecosystem enthusiasm. Do not expose every tool just because the server can. Separate read actions from write actions. Validate every tool input. Require confirmation on sensitive actions. Log enough that you can reconstruct what happened when something goes wrong.

These are not nice-to-haves. They are the default operating controls for a system that lets models reach into real tools.

What a sane production posture looks like

A strong production posture treats server metadata as separate from trust, remote transport as an authorization problem, and destructive actions as workflows that deserve explicit review points.

The more a system can modify, send, purchase, or publish, the more those review boundaries matter.

Quick decision table

SituationBetter default
Remote transportUse explicit authorization and audience validation
Sensitive toolRequire confirmation and audit logs
Broad server surfaceScope capabilities to the workflow
Tool failuresValidate inputs and sanitize outputs

Practical checklist

FAQ

Is MCP security mostly a server problem?

No. It is shared across client policy, transport, server implementation, and workflow UX.

Should every risky tool require confirmation?

As a default, yes. The higher the side effects, the stronger the review step should be.

Sources and further reading

๐Ÿ”— Security Best Practices for Model Context Protocol implementations
Official MCP security guidance covering attack vectors, mitigations, and implementation best practices.
๐Ÿ”— Authorization in MCP: transport-level auth and OAuth-based flows
Official MCP authorization specification for HTTP-based transports and security alignment.
๐Ÿ”— MCP tool security considerations: validate inputs, confirm sensitive actions, log usage
Official MCP tools specification with concrete security guidance for tool execution.
๐Ÿ”— MCP architecture overview: clients, servers, resources, prompts, and transports
Official MCP architecture overview covering the protocol model and ecosystem components.
๐Ÿ”— Linux Foundation Agentics Day: MCP plus agents is now an ecosystem topic, not a niche hack
Linux Foundation event page highlighting real-world MCP and agent implementations and best practices.

Related reading

Use this inside Thinkly

If you want your AI research, comparisons, and workflow decisions to stay reusable, keep them in Thinkly instead of scattering them across chats and tabs.

th
Made with ThinklyCollect clips. Structure thinking. Share.
Try Thinkly โ†’