Moxt

Moxt is an AI-native workspace discussed in “AGI 来了？我用了一周，头皮发麻“｜对谈张昊然：Moxt 联合创始人. Zhang Haoran presents it as a place where people create, configure, and train AI Coworkers that live inside a team’s Organizational Context rather than only answering isolated prompts.

The product name comes from “more context.” Moxt’s core design claim is that AI work improves when the workspace itself is easy for agents to read and write: documents use Markdown, tables can use CSV, structured data can use JSON, visual output can use HTML, and the space is organized with a file-system-like structure. In that framing, AI-Native Workspace value comes from reducing context loss between human work artifacts and agent execution.

The episode describes Moxt as both a product and an internal operating system for the team. The team uses it to revise documents from meeting recordings, connect data warehouses for analysis, replace some Jira-like project management with Generated Work Interfaces, and let AI coworkers such as Momo or role-specific agents monitor progress, propose sales plans, critique priorities, and participate in comments.

Moxt’s stated values matter to the product definition. Zhang Haoran says the team wants to amplify people, not sell AI as a low-cost human replacement, and treats privacy/security as another baseline. The episode describes an Agents MD document as a high-authority instruction layer that agents follow like a workspace constitution.

Connections

• Zhang Haoran — co-founder and episode guest explaining Moxt’s thesis.
• Momo — default personal assistant inside Moxt.
• AI-Native Workspace — product category Moxt is trying to define.
• Organizational Context — shared context layer Moxt tries to make agent-readable.
• AI Coworkers — role-based agents that live and work inside the space.
• Generated Work Interfaces — Moxt pattern of generating boards, dashboards, and visualizations from context.
• Context Engineering, Agentic Workflow, and Persistent Agent Memory — broader agent mechanisms Moxt depends on.
• Human-Agent Collaboration, Digital Employees, and AI Organization Design — work and organization implications raised by the product.
• Slack and Agent-Facing Interfaces — current IM friction and callable-surface questions the episode surfaces.