local-counsel-manager-scott-margetts
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a mandatory identification validation step ('hard gate') that prevents the generation of formal legal records until the user confirms specific client and matter identifiers, ensuring proper context and preventing data leakage across matters.
- [SAFE]: It features a dedicated input classification logic that requires users to tag external data (e.g., using
[FROM LC]). This protocol helps the agent distinguish between trusted user commands and potentially untrusted third-party correspondence, significantly reducing the surface for indirect prompt injection. - [SAFE]: Integration with Microsoft 365 services such as Outlook, SharePoint, and Teams is handled through the standard Model Context Protocol (MCP) and relies on platform-level authorizations to access and manage matter-related data.
- [SAFE]: External links and dependencies are limited to official GitHub repositories associated with the legal project management toolset for documentation purposes.
Audit Metadata