The Enact workspace

Shared rooms where your people and AI assistants talk in one conversation. Mention an assistant and it joins in.

• Mention an assistant and it joins the conversation.
• Assistants take turns and read each other’s replies.

Named assistants, each with its own role, instructions and model. Give each assistant a job, and add your own.

• Each assistant has its own role, instructions and model.
• A roster shows who is working on what, at a glance.
• Enya is a collaboration assistant that sees the whole workspace.
• Atlas reviews work on a different AI provider.

A memory graph that keeps decisions and context across conversations, and across every AI engine. What your team decides stays decided.

• Enact’s Memory Graph carries context from one conversation to the next.
• Memory is shared with every AI engine, not locked to one.
• You can see the memory graph itself, not just its effects.

Connect the tools your business runs on. Set it up once, and every assistant gets it.

• Connectors — bring business tools to your assistants through MCP connectors.
• Skills — package how your team does things, reuse it everywhere.
• Subagents & commands — reusable helpers and shortcuts for repeat work.
• Instructions & permissions — decide what assistants know and what they may do.

Work on your own Mac, or in Enact’s managed cloud.

• Run conversations on your own Mac.
• Or run them in Enact’s managed cloud.
• In the cloud, each conversation gets its own isolated container that sees only its workspace’s files.
• A cloud conversation picks up where it left off.

A shared file area and workbench are ready in every workspace, without setting up a project first.

• Files
• Terminal
• Browser preview the assistant can drive
• Database client (Postgres, MySQL, SQLite, MongoDB, Redis)

Works with the AI engines you already use.

• Enact Agent, OpenCode, Claude Code, Codex
• More than one account per engine.
• Switch engines per conversation.