# Deploying custom osquery extensions in Fleet ### Links to article series: - Part 1: Deploying custom osquery extensions in Fleet - Part 2: [Deploying custom osquery extensions in Fleet: A step-by-step guide](https://fleetdm.com/guides/deploying-custom-osquery-extensions-in-fleet-a-step-by-step-guide) One of the advantages of adopting open-source solutions is their extensibility. Another is the ability to customize them to meet your exact requirements. When it comes to Fleet and `osquery`, this flexibility becomes particularly powerful when deploying custom extensions across your infrastructure. ## Why custom osquery extensions matter Custom osquery extensions allow you to gather specific telemetry data that standard `osquery` or Fleet tables might not provide. Deploying custom extensions also allows you to query the data directly using simple `SQL` syntax, rather than dealing with the hassle of setting up [Automatic Table Construction (ATC)](https://www.linkedin.com/pulse/from-data-gaps-actions-auto-table-construction-atc-fleet-houchins-l9kqc/) or parsing files directly using tables like [`file`](https://fleetdm.com/tables/file#apple), [`file_lines`](https://fleetdm.com/tables/file_lines#apple), [`parse_json`](https://fleetdm.com/tables/parse_json#apple), and so on. Custom extensions give you the power to tailor your endpoint visibility to your organization's unique needs. ## Balancing functionality and infrastructure overhead Fleet supports deploying custom extensions through a custom [TUF (The Update Framework)](https://theupdateframework.io/) server, as documented in the [agent configuration guide](https://fleetdm.com/docs/configuration/agent-configuration). While this approach provides centralized management and security benefits, it also introduces additional infrastructure complexity. The TUF custom extension workflow was born out of necessitiy - before Fleet had built-in custom package deployment options. If you're looking to minimize infrastructure management overhead, you might wonder: "What if I could install custom extensions directly on hosts and have Fleet automatically pick them up?" The good news is: you can. ## Automating deployment with policy-based remediation Manual deployment doesn't scale well across large fleets. Here's where Fleet's policy-based automation becomes invaluable. ### The automation strategy - **Create a Detection Policy:** Write a policy that checks for the existence of your custom extension - **Leverage Policy Failure:** Since the extension doesn't exist initially, hosts will fail this policy - **Implement Automated Remediation:** Use Fleet's automated policy remediation to trigger installation ### Benefits of this approach - **Reduced infrastructure complexity:** No need to maintain a separate TUF server - **Familiar deployment methods:** Use existing package management and scripting - **Automated scale:** Policy-based remediation handles deployment across your entire fleet ## Putting it into production While TUF servers provide enterprise-grade extension management, direct deployment offers a pragmatic alternative for organizations seeking to minimize infrastructure overhead. By combining direct filesystem deployment with Fleet's policy-based automation, you can achieve scalable custom extension deployment without the complexity of additional server infrastructure. This approach exemplifies the flexibility that makes open source solutions so powerful — giving you the freedom to implement the deployment strategy that best fits your organization's needs and constraints. [Talk to Fleet](https://fleetdm.com/contact) to learn more! To get the step-by-step guide for devlievering custom osquery extensions with example custom extensions from Fleet's internal deployment, see the [next article](https://fleetdm.com/guides/deploying-custom-osquery-extensions-in-fleet-a-step-by-step-guide) in this series. About the author: Allen Houchins is Head of IT & Solutions Consulting at Fleet Device Management.