The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates querying and processing records from external ServiceNow tables, which creates an ingestion surface for untrusted data.
Ingestion points: Data retrieved through the snow_query_table and snow_execute_script tools (SKILL.md).
Boundary markers: The provided patterns do not specify delimiters or boundary markers to isolate record data from agent instructions.
Capability inventory: The skill utilizes snow_execute_script for server-side JavaScript execution and snow_query_table for data retrieval.
Sanitization: While the skill documentation mentions "Never trust user input," it does not provide specific implementation details for sanitizing or escaping the retrieved data.
[DYNAMIC_EXECUTION]: The skill provides patterns for creating and running server-side JavaScript through the snow_execute_script tool, which executes code dynamically on the ServiceNow platform.
[PRIVILEGE_ESCALATION]: The skill includes usage patterns for setWorkflow(false), a ServiceNow API call that allows database operations to bypass business rules, engines, and auditing protocols, potentially enabling the bypass of established security controls.

gliderecord-patterns