What Multiplayer OS Is

Multiplayer OS is the shared company workspace for humans and AI. It gives a company one governed place for wiki knowledge, agent skills, source context, approval decisions, and audit history.

The local app and CLI let existing tools report useful moments without asking agents to handle platform credentials. The managed platform keeps the workspace, permissions, sync, and history consistent across the team.

Company wiki

Durable project, decision, people, system, and meeting context that humans can inspect and agents can cite.

Agents

Local or hosted AI workers with scoped access to wiki paths, tools, skills, and external actions.

Skills

Reusable instructions and workflows that can be published by the workspace and installed into tools such as Codex, Claude, Cursor, Hermes, and others.

Sources

Connected work systems such as GitHub, Linear, Slack, Google Drive, Notion, Granola, and local scripts.

Governance

Roles, policies, approval gates, source references, risk tiers, and audit trails around every important change.

• Single-player AI tools create useful work, but the context and decisions stay trapped in individual chats, laptops, and vendor workspaces.
• AI code generation is moving faster than review, test evidence, permission policy, and release governance.
• Meeting notes, Slack threads, tickets, repos, and customer context need synthesis before agents should use them as company knowledge.
• Companies need a way for AI tools to propose changes without silently mutating the shared source of truth.
• Teams need shared skills and policies so each agent is not relearning the company from scratch.

• Not another standalone chat surface.
• Not a blind crawler that uploads local files or transcripts.
• Not a vendor-owned memory silo.
• Not a way for agents to bypass human review.
• Not a replacement for existing tools; it coordinates the tools the team already uses.