OpenCompanion Launches as Open-Source Desktop Manager for AI Coding CLIs

OpenCompanion has launched as an open-source desktop application designed to streamline the management of multiple AI coding CLI sessions. The tool addresses the friction of running agents like Claude Code, Codex CLI, and OpenCode in parallel, where each session frequently pauses for permission requests that users must manually check in individual terminals.

Core Features

The application provides a unified dashboard to start, watch, and answer AI coding CLIs:

• Session Management: Starts CLIs in real terminals or headless mode, listing all active sessions on one screen with status indicators.
• Remote Approval: Notifies users when a session is waiting for input. For Claude Code, it provides Approve and Deny buttons accessible from both the desktop and a mobile device on the same Wi-Fi network.
• Automation: Converts plain requests into individual cards per session, allowing users to review prompts before running them. It also supports scheduled automations for repetitive tasks.
• Developer Context: Offers split-shell views beside each session to run tests, browse folders, and read uncommitted diffs.

Privacy and Platform Support

OpenCompanion operates entirely locally with no account, no cloud server, and no telemetry. Each CLI continues to use the user's own login credentials and quota.

• Availability: Installers for Windows, Linux, and macOS are available via GitHub Releases and the project website. The software is MIT licensed.
• Current Limitations: Installers are not yet code-signed, requiring a one-time permission allowance on first start. The mobile connection uses plain HTTP on the local network, so remote access requires a VPN like Tailscale.