The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from ServiceNow tables that are then processed by the agent.
Ingestion points: Data is retrieved from the sn_risk_identification (e.g., short_description), sn_grc_risk (e.g., short_description), sn_grc_profile, and sn_grc_issue tables as described in the Procedure section of SKILL.md.
Boundary markers: The instructions lack explicit boundary markers or directives for the agent to ignore or isolate instructions potentially embedded within the records retrieved from ServiceNow.
Capability inventory: The skill allows creating and updating records via SN-Create-Record, SN-Update-Record, and SN-Add-Work-Notes. It also lists Bash as a native tool in the YAML frontmatter of SKILL.md.
Sanitization: There are no instructions for sanitizing or validating input from ServiceNow before the agent acts upon it or generates scoring/rationales.

risk-identification