💡 摘要
Ralph TUI 是一个终端用户界面,用于协调 AI 编码代理自主完成任务。
🎯 适合人群
🤖 AI 吐槽: “README 提示潜在风险,如远程访问和令牌管理,如果处理不当可能会暴露敏感数据。实施安全的令牌存储和定期轮换可以减轻这些风险。”
README 提示潜在风险,如远程访问和令牌管理,如果处理不当可能会暴露敏感数据。实施安全的令牌存储和定期轮换可以减轻这些风险。
Ralph TUI
AI Agent Loop Orchestrator - A terminal UI for orchestrating AI coding agents to work through task lists autonomously.
Ralph TUI connects your AI coding assistant (Claude Code, OpenCode, Factory Droid, Gemini CLI, Codex, Kiro CLI) to your task tracker and runs them in an autonomous loop, completing tasks one-by-one with intelligent selection, error handling, and full visibility.

Quick Start
# Install bun install -g ralph-tui # Setup your project cd your-project ralph-tui setup # Create a PRD with AI assistance ralph-tui create-prd --chat # Run Ralph! ralph-tui run --prd ./prd.json
That's it! Ralph will work through your tasks autonomously.
Documentation
ralph-tui.com - Full documentation, guides, and examples.
Quick Links
- Quick Start Guide - Get running in 2 minutes
- Installation - All installation options
- CLI Reference - Complete command reference
- Configuration - Customize Ralph for your workflow
- Troubleshooting - Common issues and solutions
How It Works
┌─────────────────────────────────────────────────────────────────┐
│ │
│ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │
│ │ 1. SELECT │────▶│ 2. BUILD │────▶│ 3. EXECUTE │ │
│ │ TASK │ │ PROMPT │ │ AGENT │ │
│ └──────────────┘ └──────────────┘ └──────────────┘ │
│ ▲ │ │
│ │ ▼ │
│ ┌──────────────┐ ┌──────────────┐ │
│ │ 5. NEXT │◀────────────────────────│ 4. DETECT │ │
│ │ TASK │ │ COMPLETION │ │
│ └──────────────┘ └──────────────┘ │
│ │
└─────────────────────────────────────────────────────────────────┘
Ralph selects the highest-priority task, builds a prompt, executes your AI agent, detects completion, and repeats until all tasks are done.
Features
- Task Trackers: prd.json (simple), Beads (git-backed with dependencies)
- AI Agents: Claude Code, OpenCode, Factory Droid, Gemini CLI, Codex, Kiro CLI
- Session Persistence: Pause anytime, resume later, survive crashes
- Real-time TUI: Watch agent output, control execution with keyboard shortcuts
- Subagent Tracing: See nested agent calls in real-time
- Cross-iteration Context: Automatic progress tracking between tasks
- Flexible Skills: Use PRD/task skills directly in your agent or via the TUI
- Remote Instances: Monitor and control ralph-tui running on multiple machines from a single TUI
CLI Commands
| Command | Description |
|---------|-------------|
| ralph-tui | Launch the interactive TUI |
| ralph-tui run [options] | Start Ralph execution |
| ralph-tui resume | Resume an interrupted session |
| ralph-tui status | Check session status |
| ralph-tui logs | View iteration output logs |
| ralph-tui setup | Run interactive project setup |
| ralph-tui create-prd | Create a new PRD interactively |
| ralph-tui convert | Convert PRD to tracker format |
| ralph-tui config show | Display merged configuration |
| ralph-tui template show | Display current prompt template |
| ralph-tui plugins agents | List available agent plugins |
| ralph-tui plugins trackers | List available tracker plugins |
| ralph-tui run --listen | Run with remote listener enabled |
| ralph-tui remote <cmd> | Manage remote server connections |
Common Options
# Run with a PRD file ralph-tui run --prd ./prd.json # Run with a Beads epic ralph-tui run --epic my-epic-id # Override agent or model ralph-tui run --agent claude --model sonnet ralph-tui run --agent opencode --model anthropic/claude-3-5-sonnet # Limit iterations ralph-tui run --iterations 5 # Run headless (no TUI) ralph-tui run --headless # Run agent in isolated sandbox (bwrap on Linux, sandbox-exec on macOS) # Requires bwrap to be installed and on PATH (Linux) or uses built-in sandbox-exec (macOS) ralph-tui run --sandbox
Create PRD Options
# Create a PRD with AI assistance (default chat mode) ralph-tui create-prd ralph-tui prime # Alias # Use a custom PRD skill from skills_dir ralph-tui create-prd --prd-skill my-custom-skill # Override agent ralph-tui create-prd --agent claude # Output to custom directory ralph-tui create-prd --output ./docs
TUI Keyboard Shortcuts
| Key | Action |
|-----|--------|
| s | Start execution |
| p | Pause/Resume |
| d | Toggle dashboard |
| T | Toggle subagent tree panel (Shift+T) |
| t | Cycle subagent detail level |
| o | Cycle right panel views |
| , | Open settings (local tab only) |
| C | Open read-only config viewer (Shift+C, works on local and remote tabs) |
| q | Quit |
| ? | Show help |
| 1-9 | Switch to tab 1-9 (remote instances) |
| [ / ] | Previous/Next tab |
| a | Add new remote instance |
| e | Edit current remote (when viewing remote tab) |
| x | Delete current remote (when viewing remote tab) |
Dashboard (d key): Toggle a status panel showing:
- Current execution status and active task
- Agent name and model (e.g.,
claude-code,anthropic/claude-sonnet) - Tracker source (e.g.,
prd,beads) - Git branch with dirty indicator (repo:branch*)
- Sandbox status (🔒 enabled, 🔓 disabled) with mode
- Auto-commit setting (✓ auto, ✗ manual)
- Remote connection info (when viewing remote tabs)
See the full CLI reference for all options.
Using Skills Directly in Your Agent
Install ralph-tui skills to your agent using add-skill:
# Install all skills to all detected agents globally bunx add-skill subsy/ralph-tui --all # Install to a specific agent bunx add-skill subsy/ralph-tui -a claude-code -g -y # Or use the ralph-tui wrapper (maps agent IDs automatically) ralph-tui skills install ralph-tui skills install --agent claude
Use these slash commands in your agent:
/ralph-tui-prd # Create a PRD interactively /ralph-tui-create-json # Convert PRD to prd.json /ralph-tui-create-beads # Convert PRD to Beads issues
This lets you create PRDs while referencing source files (@filename) and using your full conversation context—then use ralph-tui run for autonomous execution.
Custom Skills Directory
You can configure a custom skills_dir in your config file to use custom PRD skills:
# In .ralph-tui/config.toml or ~/.config/ralph-tui/config.toml skills_dir = "/path/to/my-skills" # Then use custom skills ralph-tui create-prd --prd-skill my-custom-skill
Skills must be folders inside skills_dir containing a SKILL.md file.
Remote Instance Management
Control multiple ralph-tui instances running on different machines (VPS servers, CI/CD environments, development boxes) from a single TUI.
┌─────────────────────────────────────────────────────────────────┐
│ LOCAL [1]│ ● prod [2]│ ◐ staging [3]│ ○ dev [4]│ + │
├─────────────────────────────────────────────────────────────────┤
│ │
│ Your local TUI can connect to and control remote instances │
│ │
└─────────────────────────────────────────────────────────────────┘
Quick Start: Remote Control
On the remote machine (server):
# Start ralph with remote listener enabled ralph-tui run --listen --prd ./prd.json # First run generates a secure token - save it! # ═══════════════════════════════════════════════════════════════ # Remote Listener Enabled # ═══════════════════════════════════════════════════════════════ # Port: 7890 # New server token generated: # OGQwNTcxMjM0NTY3ODkwYWJjZGVmMDEyMzQ1Njc4OQ # ⚠️ Save this token securely - it won't be shown again! # ═══════════════════════════════════════════════════════════════
On your local machine (client):
# Add the remote server ralph-tui remote add prod server.example.com:7890 --token OGQwNTcxMjM0NTY3... # Test the connection ralph-tui remote test prod # Launch TUI - you'll see tabs for local + remote instances ralph-tui
Remote Listener Commands
Recommended: Use run --listen (runs engine with remote access):
# Start with remote listener on default port (7890) ralph-tui run --listen --prd ./prd.json # Start with custom port ralph-tui run --listen --listen-port 8080 --epic my-epic
Token management:
# Rotate authentication token (invalidates old token immediately) ralph-tui run --listen --rotate-token --prd ./prd.json # View remote listener options ralph-tui run --help
Remote Configuration Commands
# Add a remote server ralph-tui remote add <alias> <host:port> --token <token> # List all
优点
- 支持多种 AI 代理
- 实时任务跟踪和可视化
- 中断任务的会话持久性
- 与自定义技能的灵活集成
缺点
- 需要初始设置和配置
- 新用户可能有学习曲线
- 依赖外部 AI 代理
- 仅限于终端界面
相关技能
免责声明:本内容来源于 GitHub 开源项目,仅供展示和评分分析使用。
版权归原作者所有 subsy.
