The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill uses the snow_execute_script tool to execute arbitrary server-side JavaScript on a ServiceNow instance. This enables the agent to generate and run scripts dynamically based on templates provided in the skill body, which could be abused to perform unauthorized operations if the agent is misled.\n- [DATA_EXFILTRATION]: Code examples in the skill explicitly retrieve and process sensitive employee data, including annual salary information (u_annual_salary) and organizational roles. These sensitive details are ingested into the agent's context during processes like document generation, increasing the risk of exposure.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources that can be influenced by users, creating a risk of indirect prompt injection.\n
Ingestion points: Data is pulled from the sn_hr_core_case table and the ritm.variables object from Catalog Item requests.\n
Boundary markers: The skill does not employ specific boundary markers or instructions to ignore embedded commands within the processed data.\n
Capability inventory: The skill possesses high-privilege tools such as snow_execute_script, snow_query_table, and snow_artifact_manage which can modify system state if malicious instructions are processed.\n
Sanitization: The skill does not perform validation or sanitization on external data before using it in script logic or document templates.

hr-service-delivery