threat-intel-platform-attack
Fail
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: CRITICALDATA_EXFILTRATIONPERSISTENCEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [DATA_EXFILTRATION]: The skill provides specific payloads to reconfigure platform enrichment services to send sensitive threat data to an external collection server (e.g.,
https://attacker.example.com/collect). - [PERSISTENCE]: Instructions are included for establishing long-term access by creating unauthorized administrative users and registering attacker-controlled synchronization servers that facilitate continuous data flow from the attacker to the victim platform.
- [DYNAMIC_EXECUTION]: The skill makes extensive use of Python heredocs (
python3 << EOF) to dynamically generate and execute scripts for manipulating STIX/TAXII feeds and interacting with platform APIs. - [INDIRECT_PROMPT_INJECTION]: A core component of the skill is the injection of false-positive Indicators of Compromise (IOCs) into sharing feeds, which acts as an indirect injection vector against downstream security systems that automatically ingest and act on these feeds.
- [COMMAND_EXECUTION]: Includes administrative commands intended to delete audit logs and remove evidence of unauthorized activity on MISP and OpenCTI instances.
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of multiple Python packages, such as
stix2,taxii2-client, andfeedgenerator, which are used to construct and transmit malicious threat intelligence data.
Recommendations
- Contains 4 malicious URL(s) - DO NOT USE
Audit Metadata