detection-engineering-coverage-evaluation

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests content from external URLs and raw text inputs to drive a detection engineering workflow. This provides a surface for indirect prompt injection, where an attacker-controlled webpage could attempt to manipulate the resulting security rules or agent behavior.
  • Ingestion points: Step 1 ('Extract Threat Intelligence') accepts arbitrary URLs and processes the retrieved content.
  • Boundary markers: The skill includes an explicit 'Check for Prompt Injection' step (Step 1.3) with patterns to halt execution if malicious instructions are detected in the input.
  • Capability inventory: The skill has the ability to read existing SIEM rules using get_rule and create new rules using create_rule in the Google SecOps environment.
  • Sanitization: Instructions require the decomposition of HTML elements (removing scripts and styles) and cleaning of UI boilerplate before processing the text.
  • Mitigation: A critical safety control is present in Step 8, which requires explicit user review and approval for every rule before it is created in the production environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 05:40 AM
Security Audit — agent-trust-hub — detection-engineering-coverage-evaluation