codexia
Agent Workstation for Codex CLI + Claude Code — with task scheduler, git worktree & remote control
claude mcp add --transport stdio milisp-codexia node src/main.js
How to use
Codexia is a powerful GUI and toolkit that integrates Codex CLI and Claude code workflows into a single desktop experience. It offers project and session management, a built-in multi-file viewer, prompt notepad, remote browser control, and a management layer for MCP servers. With Codexia you can navigate Codex projects, review session histories, run sandboxed code, and toggle built-in GPT-powered features like code search and agent-based assistance. The MCP server aspect enables one-click installation and management of modular MCP servers (such as Codexia itself and other tools) from a centralized marketplace or via direct linking, so you can extend the IDE with additional capabilities as needed.
To use Codexia’s capabilities, launch the app to access the Project & Session Manager, which shows Codex CLI projects and Claude code projects. Use the built-in one-click PDFs, CSV/XLSX previews, and the file-tree to preview and navigate assets. The Prompt Notepad lets you chat with agents or scripts, and the AI-driven tools (including multiple providers like OpenAI and Claude) power code completion, session histories, and data insights. If you manage MCPs, you can add or sync servers from the marketplace or via mcp-linker to extend Codexia with new tooling and workflows.
You can customize task execution with sandboxed runtimes, approval workflows, and per-session isolation, ensuring secure testing of code and commands. The UI maintains persistent state across sessions, provides a web preview (e.g., Next.js served locally), and supports themes and accents for a tailored experience.
How to install
Prerequisites:
- Git must be installed on your system
- Codex CLI: Install from https://github.com/openai/codex
- Claude Code CLI: Install from https://claude.ai/code
- Optional: Node.js runtime if you plan to run the MCP server via Node (as configured in this guide)
Installation steps:
- Clone or download the Codexia repository to your local machine.
- If the project is Node-based, ensure Node.js is installed (e.g., nvm use, then npm install). If it is provided as a desktop package, install via the recommended method (homebrew on macOS, or the provided installer).
- Install dependencies and build steps as described in the repository’s documentation. Typical steps may include:
# Example (adjust if your project uses yarn or a different package manager)
npm install
npm run build
- Start the MCP server using the configuration described in mcp_config. If you’re following this guide, the server can be started via the command defined in the MCP configuration (for example, node src/main.js or an npm script).
- Open the Codexia application, then navigate to MCP Server Management to verify the codexia server is online and ready to manage additional MCPs.
Notes:
- If you encounter port collisions, adjust the server’s port in your environment or configuration file.
- Ensure Codex CLI and Claude Code CLI are accessible in your PATH as required by Codexia.
Additional notes
Tips and common issues:
- Ensure Codex CLI and Claude Code CLI are installed and properly authenticated for access to necessary models.
- When enabling multiple MCPs, use the marketplace or mcp-linker to import additional servers without breaking the primary Codexia setup.
- If the GUI cannot load previews, check file permissions for ~/.codex/config.toml and ~/.codex/sessions to ensure Codexia can read and write.
- For complex projects, leverage the session history and filtering to quickly find relevant conversations and tasks.
- If you run into environment variable issues, you can set variables in the mcp_config under the env field for the Codexia server (e.g., API_KEYS, model endpoints).
- Review AGENTS.md within Codexia for built-in agent editing, live previews, and markdown tooling to customize agent-assisted workflows.
Related MCP Servers
superset
IDE for the AI Agents Era - Run an army of Claude Code, Codex, etc. on your machine
sre
The SmythOS Runtime Environment (SRE) is an open-source, cloud-native runtime for agentic AI. Secure, modular, and production-ready, it lets developers build, run, and manage intelligent agents across local, cloud, and edge environments.
DeepMCPAgent
Model-agnostic plug-n-play LangChain/LangGraph agents powered entirely by MCP tools over HTTP/SSE.
claude-codex-settings
My personal Claude Code and OpenAI Codex setup with battle-tested skills, commands, hooks, agents and MCP servers that I use daily.
OpenContext
A personal context store for AI agents and assistants—reuse your existing coding agent CLI (Codex/Claude/OpenCode) with built‑in Skills/tools and a desktop GUI to capture, search, and reuse project knowledge across agents and repos.
deepcontext
DeepContext is an MCP server that adds symbol-aware semantic search to Claude Code, Codex CLI, and other agents for faster, smarter context on large codebases.