Get started with Meandr.

Point an agent at Meandr, connect the tools it should reach, and set the rules. Three short guides take you from zero to a governed setup.

Quickstart

Connect your first agent: one URL and a bearer token in your MCP config. No SDK, no code changes.

Read the guide

Connect a tool server

Add an MCP server, hand Meandr its credentials — API keys, custom headers, basic auth, OAuth, mTLS or AWS SigV4 — and curate what agents see.

Read the guide

Your first policy

Write a rule: allow, deny, or require approval — matched on the agent, the tool, and even the argument values.

Read the guide

Go deeper

Past the setup steps: short reads on the ideas Meandr is built on and the wider MCP landscape — concepts, security and vendor-neutral references you can hand a teammate.

What is an MCP gateway?

What MCP is, why you would put a gateway in front of it, and exactly what Meandr terminates and forwards.

Read the guide

Argument-level policy vs allowlists

Why tool-level allowlists fall short against coarse tools, and how matching on the argument values closes the gap.

Read the guide

AWS SigV4 without a proxy

Why AWS-hosted MCP needs SigV4, why naive relays cannot sign it, and how Meandr signs the exact bytes it forwards.

Read the guide

What belongs in an audit log

What each per-call record holds, metadata by default, opt-in encrypted capture per rule, and how retention scales with your plan.

Read the guide

Approval patterns & one-time codes

How the approve action pauses a call, the ways it can reach a person, and one-time codes as the attribution model.

Read the guide

Team roles & access control

The roles model: account roles and per-project access, so each person sees only the projects they should.

Read the guide

Policy recipe library

Copy-pasteable, plain-language policies: block deletes, require approval over an amount, read-only a server, rate-limit a project.

Browse the recipes

MCP tool poisoning

A vendor-neutral field guide: what tool poisoning is, the attack classes, the public data on exposed servers, and concrete defenses.

Read the guide

Gateway vs proxy vs router

A decision framework separating a pass-through proxy from a router from a gateway, with a capability table and honest guidance on when each is enough.

Read the guide

OAuth 2.1 & PKCE for MCP

An implementer's guide to how MCP uses OAuth 2.1 and PKCE for agent identity — discovery, dynamic registration, the flow step by step, and the pitfalls.

Read the guide

OWASP Agentic Top 10

Maps the OWASP Top 10 for Agentic Applications to the MCP-layer controls that mitigate each risk — and says where a gateway does not help.

Read the guide

The MCP 2026-07-28 spec

A vendor-neutral reference to the 2026-07-28 MCP specification, read through the lens of anyone running a gateway: statelessness, header routing, auth hardening and the deprecation clock.

Read the reference

Ready when you are.

Point your agent at one endpoint and you're governed in minutes — no code changes, no proxy to run.

Get started

Questions? Email us