engineering-threat-detection-engineer
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from SIEM logs and threat intelligence feeds to generate detection logic. It lacks explicit instructions for the agent to use boundary markers or sanitize this external content, which presents a surface for indirect prompt injection.
- Ingestion points: SIEM log data (Splunk, Sentinel, Elastic, etc.), threat intelligence feeds, and MITRE ATT&CK technique updates.
- Boundary markers: Absent in the provided instructions.
- Capability inventory: Generation of Sigma rules, SIEM-specific queries (SPL, KQL, EQL), and GitHub Actions CI/CD pipeline configurations.
- Sanitization: No explicit validation or escaping of external content is instructed before interpolation into generated code or reports.
- [COMMAND_EXECUTION]: The skill provides functional templates for shell-based operations within CI/CD pipelines, including package installation and automated deployment to security platforms.
- Evidence: The 'deploy' job in the example GitHub Action uses
curlto transmit data to a SIEM REST API. - Note: The example deployment command utilizes the
-k(insecure) flag to bypass SSL verification, which is a common but sub-optimal security practice for internal testing environments.
Audit Metadata