skills/google/skills/secops-triage/Gen Agent Trust Hub

secops-triage

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection Surface: The skill is designed to ingest and analyze untrusted data, including alert titles, detection descriptions, and raw log payloads. This naturally introduces a surface for indirect prompt injection. However, the skill explicitly mitigates this risk by including a 'Prompt Injection Defense Directive' that instructs the agent to treat all incoming telemetry strictly as data and never as instructions or executable code.
  • Authorized Privileged Operations: The skill references tools capable of modifying case metadata, closing alerts, and performing manual actions like EDR isolation. These are high-privilege operations, but they are directly aligned with the intended functionality of a Security Operations triage specialist and are documented within a structured, repeatable workflow.
  • Vendor-Specific Tooling: The instructions utilize specific tools (e.g., udm_search, get_ioc_match, execute_bulk_close_case) that are consistent with the Google Security Operations platform and the stated author (Google LLC). No suspicious or third-party external dependencies were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 10:48 AM
Security Audit — agent-trust-hub — secops-triage