skills/happy-technologies-llc/happy-platform-skills/incident-activity-summarization/Gen Agent Trust Hub
incident-activity-summarization
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user comments and work notes from ServiceNow streams which could contain malicious injection payloads.\n
- Ingestion points: Data enters the context from the
sys_journal_fieldtable as shown inSKILL.md(Step 2).\n - Boundary markers: There are no explicit boundary markers or instructions to isolate or ignore embedded instructions within the fetched activity stream data.\n
- Capability inventory: The skill possesses the
SN-Add-Work-Notescapability (Step 10) to write summaries back into the incident record.\n - Sanitization: No input sanitization or validation of the retrieved journal entries is performed before passing them to the model for summarization.
Audit Metadata