csm-suggested-steps
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from ServiceNow case fields, customer comments, and historical work notes to generate resolution guidance. This external content could contain malicious instructions designed to manipulate the agent's behavior. \n
- Ingestion points: Data is retrieved from
sn_customerservice_case,sys_journal_field, andinteractiontables using theSN-Query-TableandSN-Get-Recordtools, as well as REST APIGETrequests. \n - Boundary markers: The instructions do not specify the use of delimiters or clear separation between the retrieved data and the agent's own instructions, increasing the risk that the agent might follow embedded commands. \n
- Capability inventory: The skill possesses the
SN-Update-Recordtool for writing back to the case and theBashnative tool for executing shell commands, providing a path for potential exploitation. \n - Sanitization: The skill lacks explicit sanitization or validation logic for the retrieved text before it is processed or used in shell commands.
Audit Metadata