gnx-threat-intelligence
gnx-threat-intelligence copied to clipboard
A commercial grade threat intelligence feed thats validated and updated every half hour.

About
gnX threat intelligence feed contains a blacklist of IP addresses that have crossed a threshold indicating malicious intent and/or potential IOC [indicator of compromise] activity. Because of the criteria standards and threat list sanitiation process, we significantly reduce the potential of false positives. Can be used as a standalone blocklist or to agument your existing threat intelligence.
The list is updated every 30 minutes for rapid threat blocking.
Formats
The following formats are available:
- latest-blacklist.raw: one IP address per-line [standard format]
- latest-blacklist.sc: semi-colon separated list of IP addresses for some firewalls (e.g Fortinet).
- latest-blacklist.csv: comma separated list of IP addresses.
- latest-blacklist.cdb: CDB format list for OSSEC/Wazuh integration
- latest-blacklist.crit: Most critical of threats that significantly exceed the criteria standards below.
- latest-blacklist.low: All threats including low severity, may contain false positives, use with caution.
Source
gnX runs a micro server farm in multiple locations consisting of a variety of honey pot solutions, often built in house, which monitor for the following types of events:
- Multiple connections to uncommon publicly exposed server ports, indicating scan activity.
- Brute force login attempts against common services such as SSH, API, Basic Auth, etc.
- Injection attempts against web applications, databases and API endpoints.
- Vulnerability scans and attacks using common/known software through signature identification.
Criteria
Before an IP addresses is comitted to the list, it must have met the following criteria:
- Observed making at least five malicious attempts against a single honey pot system over the course of one hour.
- Observed making at least one malicious attempts against three separate honey pot servers over the course of one hour.
- Observed making at least fifteen attempts against a single or multiple servers over the course of 12 hours.
The critical list increases the criteria by 100%. The low list reduces the criteria by 100%.
Sanitiation
A notable amount of effort is put into reducing the possibility of false positives, including:
- Whitelists are leveraged to ensure popular web properties aren't accidentily listed.
- Any internal IP addressing (BOGONS) are removed.
- Reverse lookups are performed to gather data points which help determine whether an item should be removed.
- Addresses are routinely removed when they no longer pose a threat (e.g being shut down by ISP/Hosting provider)
Recorded data
gnX records and generates a signifcant amount of data about each potential attacker including:
- IP Address / DNS with reverse lookup/arin/whois data generated in real time.
- Attack signatures, strings, injection parameters for later utilization in our honey pots.
- Anonymously check for open ports and responses on the attackers end to identify C&C/malicious software.