incident-activity-summarization

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted data from ServiceNow journal entries (work notes and comments), creating a surface for indirect prompt injection. A malicious actor with access to ServiceNow could potentially influence the agent's behavior by embedding instructions in incident records.
  • Ingestion points: Data is ingested from the incident, sys_journal_field, and sys_audit tables as seen in SKILL.md.
  • Boundary markers: The procedure does not specify explicit boundary markers or instructions for the agent to ignore commands within the retrieved text.
  • Capability inventory: The skill can write back to the system via SN-Add-Work-Notes and has access to a Bash environment.
  • Sanitization: No specific sanitization or filtering of the processed ServiceNow data is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 01:34 PM
Security Audit — agent-trust-hub — incident-activity-summarization