Detection: Curl Execution with Percent Encoded URL

Description

The following analytic detects the execution of the curl utility where the command line includes percent-encoded characters and explicit file output options (such as -o or --output). It leverages process execution telemetry from Endpoint Detection and Response (EDR) data sources to identify curl commands that may be using URL encoding to obfuscate download locations or payload paths. This behavior is notable because percent-encoded URLs are commonly used by adversaries to evade simple string-based detections, hide malicious infrastructure, or bypass network security controls. When combined with file download behavior, this activity may indicate malware staging, payload retrieval, or secondary tool deployment. Analysts should review the decoded URL, destination host, parent process, and downloaded file to determine whether the activity is authorized or malicious. The analytic calculates the number of percent (%) characters in the curl command line and triggers when a threshold of three or more is met, indicating potential URL encoding. Adjust the threshold as needed based on your environment and tuning requirements.

Search

 1
 2| tstats `security_content_summariesonly`
 3  count min(_time) as firstTime
 4        max(_time) as lastTime
 5from datamodel=Endpoint.Processes where
 6(
 7  Processes.process_name IN ("curl.exe", "curl")
 8  OR
 9  Processes.original_file_name="curl.exe"
10)
11Processes.process IN (
12  "* --output *",
13  "* -o *" /* Covers both options since the search is case insensitive */,
14)
15Processes.process IN ("*%*")
16by Processes.action Processes.dest Processes.original_file_name
17   Processes.parent_process Processes.parent_process_exec
18   Processes.parent_process_guid Processes.parent_process_id
19   Processes.parent_process_name Processes.parent_process_path
20   Processes.process Processes.process_exec Processes.process_guid
21   Processes.process_hash Processes.process_id
22   Processes.process_integrity_level Processes.process_name
23   Processes.process_path Processes.user
24   Processes.user_id Processes.vendor_product
25
26
27| `drop_dm_object_name(Processes)`

Count the number of % characters in the process command line. Change this threshold based on your environment and tuning needs.

 1
 2| eval percent_count = mvcount(split(process, "%")) - 1
 3
 4| where percent_count >= 3
 5
 6
 7| `security_content_ctime(firstTime)`
 8
 9| `security_content_ctime(lastTime)`
10
11| `curl_execution_with_percent_encoded_url_filter`

Data Source

Name	Platform	Sourcetype	Source
CrowdStrike ProcessRollup2	Other	'crowdstrike:events:sensor'	'crowdstrike'
Sysmon EventID 1	Windows	'XmlWinEventLog'	'XmlWinEventLog:Microsoft-Windows-Sysmon/Operational'
Sysmon for Linux EventID 1	Linux	'sysmon:linux'	'Syslog:Linux-Sysmon/Operational'
Windows Event Log Security 4688	Windows	'XmlWinEventLog'	'XmlWinEventLog:Security'

Macros Used

Annotations

Default Configuration

This detection is configured by default in Splunk Enterprise Security to run with the following settings:

Implementation

The detection is based on data that originates from Endpoint Detection and Response (EDR) agents. These agents are designed to provide security-related telemetry from the endpoints where the agent is installed. To implement this search, you must ingest logs that contain the process GUID, process name, and parent process. Additionally, you must ingest complete command-line executions. These logs must be processed using the appropriate Splunk Technology Add-ons that are specific to the EDR product. The logs must also be mapped to the Processes node of the Endpoint data model. Use the Splunk Common Information Model (CIM) to normalize the field names and speed up the data modeling process.

Known False Positives

No false positives have been identified at this time.

Associated Analytic Story

• Compromised Windows Host

• Ingress Tool Transfer

• Living Off The Land

Risk Based Analytics (RBA)

Risk Message:

An instance of $parent_process_name$ spawning $process_name$ was identified on endpoint $dest$ by user $user$ with URL-encoded parameters $process$.

References

• https://github.com/nasbench/Misc-Research/blob/main/LOLBINs/Curl.md

• https://attack.mitre.org/techniques/T1027/

• https://attack.mitre.org/techniques/T1105/

• https://curl.se/docs/manpage.html

Detection Testing

Replay any dataset to Splunk Enterprise by using our replay.py tool or the UI. Alternatively you can replay a dataset into a Splunk Attack Range

Source: GitHub | Version: 1