Engineering deep dives, product updates, and perspectives on SSH security from the Mezite team.
How MezHub checks secrets: constant-time comparison where it must compare, and hash-then-look-up for local-policy operator sessions and enrolment tokens so there is no secret comparison to time at all.
Read moreWhy MezHub replaces a device's local policy outright instead of merging it with the organisation's, and how the daemon enforces that with one atomic pointer and two guards.
Read moreWhy standard HTTP/2 breaks kubectl exec over a transparent TLS MITM proxy, and how we avoid it by staying on HTTP/1.1 and splicing the upgraded stream.
Read moreChoosing best-effort append-only JSONL files over database rows for the `msh local` audit log.
Read moreHow we replaced mTLS and bearer tokens with SO_PEERCRED and LOCAL_PEERCRED over Unix domain sockets for the local daemon-to-CLI IPC.
Read moreExamining the architectural trade-offs of intercepting AI agent traffic, handling SPDY upgrades, and extracting intent from opaque protocols.
Read moreWhy traditional binary access control fails for AI agents, and how `msh local` solves this with a four-tier risk classification model.
Read moreHow we avoided reinventing the wheel by using the classifier's generated method identities directly in our policy engine, instead of building a redundant resource matching DSL.
Read moreWhy `msh local` intentionally hangs HTTP connections while waiting for human approval instead of building an async negotiation protocol, and why simplicity often beats correctness in proxy design.
Read moreHow `msh local` implements Observe Mode to record 'would-be' decisions and safely tune transparent proxy interception without disrupting developer tools.
Read moreWhy intercepting all local HTTPS traffic breaks cert-pinned apps and kills performance, and how `msh local` solves this using a selective inspect list.
Read moreKubernetes native audit logs record what hits the API server. For inspected requests in enforce mode, `msh local` records risk tiers and approval decisions. Here's why you run both.
Read moreWhy AI coding agents changed our trade-off on request inspection, how Iddio became msh local, and why inspection belongs on the workstation rather than in MezHub.
Read moreWhy Mezite's end-to-end suite runs on rootless Podman on developer machines and Docker on CI runners, and the small seam in the test harness that lets one compose file serve both.
Read moreA technical look at administrative self-lockouts, why hiding buttons in the UI is never enough, and how Mezite enforces self-modification guards on every API surface that can change a user.
Read moreA technical look at the architecture behind rejecting bespoke database and Kubernetes protocol parsing, and why that boundary — not "SSH only" — is what keeps Mezite secure and simple.
Read moreA deep dive into why string-matching your Content-Security-Policy header in tests is dangerous, and how we parse the actual SPA entry document to prevent self-inflicted outages.
Read moreA technical look at our deliberate decision to reject database and Kubernetes proxying, and where we draw the line on what Mezite proxies at all.
Read moreWhat a live cross-tenant probe taught us about the difference between per-pod compute isolation and enforceable network boundaries on Amazon EKS.
Read moreA technical exploration of why Mezite agents use reverse tunnels to connect to the control plane, eliminating inbound firewall rules and simplifying cluster deployments.
Read moreA technical look at our decision to drop CGO and use a pure Go SQLite driver, and the specific query context cancellation patterns required to keep the system robust.
Read moreWhy we mandate ReadHeaderTimeout on all http.Server instantiations in Mezite, and how a one-line oversight can expose your infrastructure to catastrophic resource exhaustion.
Read moreA deep dive into Mezite's audit logging architecture. Discover how we decouple database writes from the fast path to guarantee low latency and high throughput.
Read moreA technical deep dive into how we use TLS ALPN to serve HTTPS, gRPC, reverse tunnels, and SSH all over a single port, simplifying deployment and firewall rules.
Read moreA technical look at our decision to standardize on ECDSA P-256 for CA keypairs, focusing on Go's crypto implementation, entropy handling, and ubiquitous compatibility.
Read moreA technical look at why proxy-based session recording is a flawed architecture for modern infrastructure, and how we implemented node-mode recording at the PTY level in Go.
Read moreToday we are launching Mezite, a self-hosted platform that replaces static SSH keys with certificate-based authentication. Single signed binary, closed source, source-available under enterprise license.
Read moreYour SSH access platform holds the keys to every server in your infrastructure. Here is why it should run on your hardware, in your network, under your control.
Read moreSSH keys have been the default for decades. They are also one of the biggest unmanaged attack surfaces in most organizations. Here is how certificates fix this.
Read more