missed-meeting-email
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to research company websites and public sources to establish context when local project files are missing.
- Ingestion points: The skill fetches external websites and user-provided URLs to determine positioning, offers, and voice (defined in
SKILL.md). - Boundary markers: Safety instructions are provided in
references/house-rules.md(Rule 7), which mandates that the agent must treat all fetched content as data rather than instructions. - Capability inventory: The skill possesses the ability to perform network fetches for research and to read/write to the local
.agents/product-context.mdfile to persist research findings. - Sanitization: The instructions explicitly require the agent to quote any embedded instructions found in data rather than acting on them.
Audit Metadata