</>MCP Agents Market
Agent

ralph-tui

by subsy2.4kTypeScriptUpdated 2026-09-01

Claude CodeCursorCodex

Ralph TUI is an AI agent loop orchestrator that autonomously manages AI coding assistants through structured task lists. It connects popular AI coding agents (Claude Code, OpenCode, Factory Droid, Cursor CLI, Gemini CLI, Codex, Kiro CLI) to task trackers, then executes tasks one-by-one in an autonomous loop with intelligent selection, error handling, and session persistence. Developers use it to transform PRDs or task lists into automated development sessions that can run unattended, pause and resume across crashes, and scale across multiple remote machines.

Key Features

Autonomous task execution loop with intelligent task selection and completion detection
Supports 7 AI coding agents: Claude Code, OpenCode, Factory Droid, Cursor CLI, Gemini CLI, Codex, and Kiro CLI
Session persistence that survives crashes with pause/resume capabilities
Real-time terminal UI with agent output visibility and keyboard shortcuts for control
Remote instance management to monitor and control ralph-tui running on multiple machines from a single TUI
Multiple task trackers: prd.json (simple) and Beads (git-backed with dependencies)
Subagent tracing to visualize nested agent calls in real-time
Sandbox execution mode using bwrap (Linux) or sandbox-exec (macOS) for isolated agent runs

Use Cases

  • 01Autonomously work through feature development from a PRD or epic across multiple coding sessions
  • 02Run unattended development tasks overnight or on remote servers with full session recovery
  • 03Manage parallel development across multiple epics with a single scheduler and merge queue
  • 04Monitor and control AI coding agents running on VPS servers or CI/CD environments from your local machine
  • 05Create PRDs interactively with AI assistance then hand off to autonomous execution
  • 06Execute coding tasks in isolated sandboxes for security-sensitive projects

Related Agents

View more

ralph-tui — FAQ

What is ralph-tui?+

Ralph TUI is an AI agent orchestrator that runs AI coding assistants (like Claude Code, OpenCode, Cursor CLI) through task lists autonomously. It selects tasks, builds prompts, executes your chosen AI agent, detects completion, and repeats until all tasks are done.

How do I install ralph-tui?+

Install via Bun with 'bun install -g ralph-tui', then run 'ralph-tui setup' in your project directory to configure it. You can then create a PRD with 'ralph-tui create-prd --chat' and start execution with 'ralph-tui run --prd ./prd.json'.

Which AI coding agents does ralph-tui work with?+

Ralph TUI supports Claude Code, OpenCode, Factory Droid, Cursor CLI, Gemini CLI, Codex, and Kiro CLI. You can switch agents using the --agent flag or through the TUI's agent picker.

Do I need API keys to use ralph-tui?+

You need API keys or authentication for whichever AI coding agent you choose to use with ralph-tui. The orchestrator itself doesn't require separate API keys, but your selected agent (Claude Code, OpenCode, etc.) will need its own credentials configured.

Is ralph-tui free?+

Yes, ralph-tui is open source under the MIT license and free to use. However, you'll incur costs from the AI coding agent services you connect it to (such as Claude API usage).

Can I control ralph-tui running on remote servers?+

Yes, ralph-tui supports remote instance management. Start the remote with 'ralph-tui run --listen', then add it to your local client with 'ralph-tui remote add' using the generated token. You can then monitor and control multiple remote instances from a single TUI.

How do I install ralph-tui?+

Open the source repository on GitHub and follow its README. ralph-tui is a agent — MCP Agents Market links you directly to the official repo.

Is ralph-tui free?+

ralph-tui is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches