scheduled-briefing
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from potentially untrusted external sources, such as email inboxes and calendar entries, which could contain hidden instructions designed to manipulate the agent.
- Ingestion points: Ingests data from inbox items, calendar entries, and assignment tracker notes as described in
SKILL.md. - Boundary markers: The skill mandates the use of
<EXTERNAL_DATA source="...">tags to isolate retrieved material. - Capability inventory: The skill utilizes read access to user productivity tools and can output briefings via a notifier or file.
- Sanitization: The instructions include an 'Untrusted content boundary' section that directs the agent to ignore embedded requests, validate all URLs, and require explicit user authorization for record changes.
Audit Metadata