The Agent Skills Directory

[SAFE]: The skill facilitates legitimate ServiceNow development tasks, including the creation of Business Rules, Client Scripts, and Script Includes. All procedures align with standard ServiceNow developer workflows.
[CREDENTIALS_UNSAFE]: The skill correctly avoids hardcoding sensitive information. It uses environment variable placeholders ($SN_USER, $SN_PASS, $SN_INSTANCE) for authentication, which is the recommended practice for managing credentials in scripts.
[EXTERNAL_DOWNLOADS]: The skill uses curl to interact with ServiceNow's REST API. These operations target the user's own ServiceNow instance and are used solely for record creation and querying as part of the intended functionality.
[INDIRECT_PROMPT_INJECTION]: While the skill involves generating executable code from user requirements, this is the primary purpose of the tool. The risk is minimized by the skill providing specific, safe boilerplate templates (e.g., IIFE wrappers for Business Rules) and following ServiceNow best practices.
Ingestion points: The skill takes user requirements as input to populate script templates in SKILL.md.
Boundary markers: None explicitly defined in the provided templates.
Capability inventory: The skill uses SN-Create-Record and curl to write to system tables (sys_script, sys_script_include, etc.).
Sanitization: The skill relies on standard ServiceNow scripting patterns which include internal platform-level protections.

code-assist