# Mapping Fleet and osquery results to the MITRE ATT&CK® framework via Splunk
## What is the MITRE ATT&CK® Framework and what role does Fleet and osquery play?
From the [MITRE ATT&CK® site](https://attack.mitre.org):"MITRE ATT&CK® is a globally-accessible
knowledge base of adversary tactics and techniques based on real-world observations." Essentially, Fleet and osquery bring the "real-world observations" to the table when you are looking to adopt the MITRE ATT&CK® framework.
While osquery is capable of gathering an extremely diverse set of data from macOS, Windows, and Linux hosts, it can be difficult understanding how to use that data. In this article we will translate the techniques and tactics defined by ATT&CK® into queries that we can execute using Fleet. Then we will provide an example aggregating our logs from Fleet into Splunk Cloud to construct a dashboard of our real-world observations. The intent of this article is to provide a starting point in correlating osquery related observations from host endpoints, using query packs already out in the wild.
First, we’re going to map osquery observations to ATT&CK®. Initially, I began combing through the current 273 tables that I have access to in my fleet instance. I quickly realized that this effort would be quite daunting. Surely, I wasn't the first one that was looking to do this, right? This is when I came across @teoseller's repo
My existing deployment is done via AWS Terraform. Please check out the [existing deployment guide](https://fleetdm.com/deploy/deploying-fleet-on-aws-with-terraform) for instructions on how to set this up. With this configuration, the logging pipeline is set to send
logs to [Amazon Kinesis Data Firehose](https://fleetdm.com/docs/using-fleet/log-destinations#amazon-kinesis-data-firehose), terminating in an AWS S3 bucket. See additional documentation around configuration of [log destinations](https://fleetdm.com/docs/using-fleet/log-destinations).
Now comes the tough part, or at least it was a bit challenging for me, since I'm no Splunk expert. We’re going to build some SPL (Search Processing Language) to translate the observations we've uncovered via osquery into search results in Splunk. After that, we can drop the search results into a dashboard or even build an alert. That being said though, if this was an alerting use case, I would recommend using the built-in Policies from Fleet to trigger alerts via webhooks. Here's what the first query looks like to get the Process Connections from our Fleet scheduled query and push it to a table in Splunk:
This is just the beginning. There are many more Tactics and Techniques that can be mapped to queries and there are plenty of queries already pre-built for MITRE ATT&CK®. The next step is to continue to build out these mappings and publish the queries that map from Fleet to MITRE ATT&CK®.
