debugging-mutations
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from external ServiceNow logs and audit trails, creating a potential surface for indirect prompt injection attacks.
- Ingestion points: Data is retrieved from the sys_audit, syslog, syslog_transaction, and sys_hub_flow_output tables via tools such as snow_inspect_mutations and snow_get_logs.
- Boundary markers: The provided documentation does not specify the implementation of boundary markers or explicit instructions for the agent to ignore potentially malicious commands embedded within the retrieved log data.
- Capability inventory: The skill allows for comprehensive inspection of system state and record mutations, including the ability to read outbound HTTP logs and flow designer execution outputs.
- Sanitization: There is no description of filtering or sanitization mechanisms to validate the contents of the ServiceNow records before they are processed by the agent.
Audit Metadata