csm-case-summarization
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an inherent attack surface because it ingests and processes untrusted data from external sources (ServiceNow tables such as
sys_journal_fieldandsys_email) which may contain adversarial instructions embedded by external users. - Ingestion points: The skill retrieves case descriptions, work notes, and email content using the
SN-Query-TableandSN-Get-Recordtools. - Boundary markers: The summary construction procedure uses clear section headers (e.g., 'TIMELINE', 'EMAILS'), but does not implement explicit delimiters or 'ignore' instructions to prevent the agent from obeying commands embedded in the retrieved data.
- Capability inventory: The agent has access to ServiceNow querying tools and Bash, which could be misused if an injection is successful.
- Sanitization: There is no evidence of sanitization or filtering of the retrieved content before it is interpolated into the summary template.
Audit Metadata