incident-sentiment

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No indicators of malicious intent, prompt injection, or unauthorized data exfiltration were found. The skill operates within the scope of a standard ServiceNow ITSM environment using provided tools.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes an attack surface for indirect prompt injection as it ingests untrusted text from external users and has capabilities to update system records. This is inherent to the sentiment analysis use case.
  • Ingestion points: Data extracted from sys_journal_field (work notes/comments) and sys_email (email bodies) in SKILL.md.
  • Boundary markers: The skill does not define explicit delimiters to separate untrusted content from instructions.
  • Capability inventory: The skill has the ability to update incident records (SN-Update-Record) and add notes (SN-Add-Work-Notes) across the script logic.
  • Sanitization: No explicit sanitization or filtering of the processed text is mentioned in the procedure.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 01:34 PM
Security Audit — agent-trust-hub — incident-sentiment