agent-session-manager
agent-session-manager is an open-source terminal session manager for AI coding agents, built with Go and Bubble Tea. It runs as a tmux TUI and provides a unified interface for launching, monitoring, and interacting with multiple agent sessions. It supports popular coding agents such as Claude Code, Codex, Gemini, Aider, Amazon Q, and OpenCode. It offers live status indicators, session grouping, and the ability to resume or fork existing agent sessions. Users can review diffs produced by agents directly in the terminal, helping to inspect code changes before accepting them. The tool also supports mobile push notifications that alert the user when an agent needs an answer or approval, allowing work to continue away from the terminal. By centralizing agent session management, it aims to reduce context switching and make parallel agent workflows more manageable for developers. The project is distributed as open-source software and is intended for terminal-centric development environments.
- Category
- AI Agents
- Pricing
- Open Source
Problem solved
Developers using multiple AI coding agents in the terminal often struggle to track sessions, switch between contexts, and notice when an agent needs input. agent-session-manager solves this by providing a single tmux-based TUI for live status, session grouping, resume/fork, diff review, and mobile push notifications, making parallel agent workflows more organized and responsive.
Use cases
- Run multiple AI coding agents in parallel and monitor their status in a single tmux TUI
- Resume or fork previous agent sessions to continue work or explore alternative approaches
- Review code diffs produced by AI agents before committing or applying changes
- Receive mobile push notifications when an agent requires input or approval
- Organize agent sessions into groups for different projects, tasks, or experiments
- Integrate with Claude Code, Codex, Gemini, Aider, Amazon Q, and OpenCode workflows
- Track live status of long-running agent tasks without constantly switching terminal windows
Audience
- Developers
- ML Engineers
- Hobbyists
- Startups
- Researchers
Pros
- Centralizes management of multiple AI coding agent sessions in one terminal UI
- Live status and session groups improve visibility across parallel agent tasks
- Supports resume and fork to continue or branch agent conversations
- Built-in diff review helps inspect AI-generated code changes before applying
- Mobile push notifications keep developers informed when agents need input
Cons
- Requires familiarity with tmux and terminal-based workflows
- Limited to supported AI coding agents (Claude Code, Codex, Gemini, Aider, Amazon Q, OpenCode)
- Mobile push notifications may require additional setup or configuration
- No native GUI; may not appeal to developers preferring graphical interfaces
- Documentation and community may be limited for a niche open-source project
Official website: https://github.com/izll/agent-session-manager
Canonical LambdaBase page: https://www.lambdabase.com/tools/agent-session-manager