daily-briefing
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from multiple external sources which could serve as a vector for indirect prompt injection attacks.
- Ingestion points: The skill reads unread customer email subjects and sender domains (Step 2d), meeting descriptions and locations (Step 2a), and searches sent-mail for unfulfilled promises (Step 3). These fields are often controlled by third parties.
- Boundary markers: No explicit delimiters or instructions (e.g., 'treat the following data as text, not instructions') are used when processing these external data points.
- Capability inventory: The skill uses
dispatch_readonlyanddispatchtools to communicate with Salesforce APIs anddisplay_widgetfor UI output. It does not have access to a shell or general-purpose network tools (curl/wget), which limits the impact of a successful injection. - Sanitization: There is no evidence of sanitization or filtering applied to external content before it is interpolated into the context for briefing generation or widget token population.
Audit Metadata