The Agent Skills Directory

[SAFE]: The skill follows established ITIL (Information Technology Infrastructure Library) patterns for Problem Management and does not contain malicious code or unauthorized data exfiltration routines.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data from ServiceNow records (incident notes, descriptions) which could contain adversarial instructions.
Ingestion points: SN-Query-Table and REST GET calls in SKILL.md are used to retrieve external incident and problem data.
Boundary markers: Absent. External data is processed and used to populate work notes and problem descriptions without explicit delimiters.
Capability inventory: The skill has access to the Bash tool and multiple ServiceNow write operations (SN-Update-Record, SN-Create-Record, SN-Close-Problem).
Sanitization: Absent. Content from ServiceNow records is used directly in analysis templates.

problem-analysis