The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill enables the agent to ingest and analyze data from ServiceNow logs and test results, which could contain instructions intended to influence the agent's behavior.
Ingestion points: External data enters the agent context through tools like SN-Get-ATF-Results (Step 7.4), SN-Get-ATF-Test-Logs (Step 7.5), and SN-Query-Table (Phase 8) in SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore instructions embedded within logs or test record fields.
Capability inventory: The skill utilizes powerful capabilities in SKILL.md, including SN-Execute-Background-Script (Step 8.3), SN-Run-ATF-Test (Step 7.2), and broad record creation/update access via SN-Create-Record and SN-Update-Record.
Sanitization: There is no evidence of content sanitization or validation for retrieved record data before it is processed by the agent.
[DYNAMIC_EXECUTION]: The skill provides patterns for executing dynamic JavaScript on the ServiceNow instance via SN-Execute-Background-Script for maintenance and reporting tasks (Step 8.3, Troubleshooting). It also creates server-side test steps containing executable scripts (Phase 3).
[COMMAND_EXECUTION]: The documentation includes shell commands for the now-sdk CLI and curl to facilitate integration with external CI/CD platforms like Jenkins and GitHub Actions (Phase 9).
[PRIVILEGE_ESCALATION]: The skill describes administrative actions such as modifying the sn_atf.runner.enabled system property via SN-Update-Record (Step 7.1) and using test steps to impersonate other users within the ServiceNow environment (Example 2).

automated-testing