extension-invite-links
Pass
Audited by Gen Agent Trust Hub on Apr 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill establishes a functional system for RSVPs where sensitive operations, such as viewing all responses or generating codes, are restricted to users with admin privileges.
- [EXTERNAL_DOWNLOADS]: All external module references (caffeineai-invite-links and caffeineai-authorization) align with the vendor's namespace and provide standard functionality for authorization and link management.
- [COMMAND_EXECUTION]: There is no evidence of unsafe command execution or shell interactions within the Motoko backend or the React frontend components.
- [PROMPT_INJECTION]: The instructions do not include any markers or patterns indicative of prompt injection or attempts to override the agent's core instructions.
Audit Metadata