Versions:

3.55.0
3.53.0
3.52.0
3.51.0
3.50.0
3.49.0
3.48.1
3.46.0
3.45.0
3.44.0
3.38.6
3.38.5
3.38.4
3.38.3
3.38.2
3.38.0
3.37.2
3.37.1
3.36.0
3.35.1
3.35.0
3.34.0
3.33.0
3.32.0
3.31.0
3.30.0
3.29.0
3.28.1
3.28.0
3.27.0
3.26.0
3.25.0
3.24.1
3.24.0
3.23.0
2.9.1
2.9.0
2.8.4
2.8.3
2.8.2
2.8.1
2.8.0
2.7.4
2.7.3
2.7.1
2.7.0
2.6.0
2.5.0

wmux is a Windows terminal multiplexer designed specifically for AI coding agents, published by openwong2kim. Its current version is 3.55.0, and it has a substantial release history comprising 48 published versions, indicating ongoing development and refinement of the software. The tool's core purpose is to enable users to run multiple AI coding agents side by side within a single Windows terminal environment, supporting popular agentic command-line tools such as Claude Code, Codex, and Gemini CLI simultaneously. Central to wmux is its built-in MCP (Model Context Protocol) server, which allows AI agents to interact with and control terminals across workspaces through dedicated MCP tools, making it well suited to workflows where automated agents need to execute commands, monitor output, and coordinate tasks without direct human intervention. Key features include split terminals for organizing multiple agent sessions within one interface, browser automation capabilities that extend agent functionality beyond the command line, and session persistence so that terminal states and running processes survive restarts or reconnections. Safety and observability are also addressed: wmux incorporates dangerous command detection to flag potentially harmful operations before execution, and it provides real-time token cost tracking so users can monitor the spending associated with their AI agent usage as it happens. In terms of category, wmux falls within developer tools, specifically terminal multiplexers and AI agent orchestration utilities, occupying a niche that bridges traditional terminal management software with modern AI-assisted development workflows. Typical use cases include running parallel coding agents on different tasks, comparing outputs from multiple AI assistants in split panes, automating browser-based interactions as part of agent-driven pipelines, and maintaining long-running agent sessions with full cost visibility. Its combination of multiplexing, MCP integration, and agent-specific safeguards positions it as a specialized terminal environment for Windows users building or supervising AI coding workflows.

Tags:

ai-agent 9

developer-tools 239

electron 267

mcp 86

multiplexer 13

terminal 254

windows 384