Lattice MCP

Lattice's MCP server connects its performance and people platform to the AI assistants employees already work in. It surfaces the written record a manager needs in one place: performance reviews, goals and their progress, feedback given and received, weekly updates, and the notes and agendas from 1:1s. An agent can gather that context for a person or a team and use it to answer a question or produce a draft. Writing is part of the design, not an afterthought. Review cycles are the heaviest recurring workload in Lattice, so the server lets an agent draft, edit and submit a review with the supporting evidence already in context, instead of asking a manager to reconstruct months of work from memory. The same context feeds meeting preparation, where the assistant assembles progress, blockers and open feedback before a 1:1. Permissions carry over unchanged. The connection uses the same login and the same visibility boundaries the person has in the Lattice interface, so an agent cannot surface reviews or feedback they could not already read, and IT administrators decide which roles are allowed to connect at all. The audience is managers, HR business partners and any employee running people workflows through an assistant.