The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses SN-Execute-Background-Script to run server-side JavaScript for analyzing case trends and extracting FAQ data. This capability allows for arbitrary script execution within the ServiceNow instance.
[DATA_EXFILTRATION]: The skill retrieves sensitive HR data including case descriptions, resolution notes, and subject person names from the sn_hr_core_case table. This information contains PII that requires careful handling to prevent unauthorized exposure.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted input from HR case fields to generate article content.
Ingestion points: Data is pulled from the sn_hr_core_case table (specifically description, short_description, and close_notes fields) in multiple steps including Step 2, Step 6, and Step 7.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the ingested case data as untrusted or to ignore any instructions found within it.
Capability inventory: The skill utilizes SN-Create-Record, SN-Update-Record, and SN-Execute-Background-Script to perform its primary tasks.
Sanitization: No programmatic sanitization or filtering is applied to the case data before it is interpolated into the prompts for article generation.

hr-kb-generation