# 31 MCP tools, topic ownership, and Schema Registry Proxy 1.0

## 31 tools for the Conduktor MCP server

The [Conduktor MCP server](https://www.conduktor.io/mcp) now has 31 tools, up from 12. An AI assistant connected to Console can read records, schemas, consumer group offsets, connectors, ACLs, stream lineage, Chargeback and the audit log.

You can ask which service accounts can write to a topic, whether a schema change is compatible, or who changed a topic's configuration last week. Each tool runs with the permissions of the token you connect with, so the assistant only sees what you can see. The first tool that changes data, `set-topic-labels`, adds labels to a topic. [See every MCP tool →](https://docs.conduktor.io/guide/conduktor-in-production/automate/mcp/tools)

## See who owns a topic and filter on its headers

For a topic owned by a Self-service application, the topic details page now shows the application and instance that own it, whether the topic is Public or Private in the Topic Catalog, and its description.

When you consume a topic, you can show any header key as a column and filter records on a header key or value. [Read about topics in Console →](https://docs.conduktor.io/guide/manage-kafka/kafka-resources/topics)

![A topic details page in Console showing the topic as Public, owned by the order-app application's order-app-dev instance, with records filtered on a header value and the region and source headers shown as columns](https://www.conduktor.io/assets/images/releases/october-2026-topic-owner-and-header-filter.png)

## Schema Registry Proxy 1.0 is generally available

[Schema Registry Proxy](https://www.conduktor.io/schema-registry-proxy) adds authentication and per-subject authorization to a Confluent-compatible schema registry. Clients point at the proxy instead of the registry, with no code changes.

Version 1.0 offers three ways to authorize requests: Self-service permissions managed in Console, your existing Kafka ACLs (a principal that can read or write a topic gets the same access to its schemas), or read-only access for dashboards and CI jobs. The last two don't need Console. It also adds HTTP Basic authentication and supports the GUID schema IDs used by Confluent Platform 8 clients. [Deploy Schema Registry Proxy →](https://docs.conduktor.io/guide/conduktor-in-production/deploy-artifacts/deploy-schema-registry-proxy)

## Manage encryption keys from one place

Gateway external connections are generally available. Declare a connection to your KMS once and reference it by name from your encryption Interceptors, instead of repeating the address and credentials in each one. This release adds Azure Key Vault, Google Cloud KMS and Fortanix DSM as KMS connections for [field-level encryption](https://www.conduktor.io/kafka-encryption), and every HashiCorp Vault auth method. [Configure external connections →](https://docs.conduktor.io/guide/conduktor-in-production/deploy-artifacts/deploy-gateway/external-connections)

## See your Gateway deployment from Console

The **Kafka Gateway** page of a [Gateway](https://www.conduktor.io/gateway) cluster now shows the Gateway version, the number of Gateway nodes and Interceptors, and the fleet state, with tabs for its listeners and nodes, Interceptors and configuration. Node details need Gateway 3.22 or later.

![The Kafka Gateway page in Console, showing the Gateway version, node and Interceptor counts, and the listeners and Gateway nodes on the Gateways tab](https://www.conduktor.io/assets/images/releases/october-2026-kafka-gateway-page.png)

## Find which stage of a request adds latency

Gateway can now time each stage of a request as a Prometheus metric: authorization, each Interceptor, the upstream queue and Kafka itself. When latency goes up, you can see which stage it came from. The metrics are off by default. [Set up request lifecycle metrics →](https://docs.conduktor.io/guide/conduktor-in-production/monitor/gateway_lifecycle_metrics)

![A Grafana dashboard of Gateway request lifecycle metrics, showing mean and max latency, throughput and the fraction of requests over 100 ms for each stage of the request](https://www.conduktor.io/assets/images/releases/october-2026-gateway-lifecycle-metrics.png)

## Also in this release

- **Default replication factor**: when you create a topic in Console, it now gets the cluster's `default.replication.factor` unless you set one under **Advanced configuration**. Previously, Console set it to the lower of three and the number of brokers.
- **Schema IDs in record headers**: Console now produces records that carry their schema GUID in a header, with Confluent Schema Registry 8.0 or later.

---

For a full list of changes, read the [complete release notes](https://docs.conduktor.io/guide/release-notes).