mikeyobrien/ralph-orchestrator
An improved implementation of the Ralph Wiggum technique for autonomous AI agent orchestration
What it solves
Ralph Orchestrator 提供一個自動化任務完成的框架,透過讓 AI 代理持續迭代直至達成目標,避免常見的「一次性」失敗——AI 助手只給出部分或錯誤答案後就停止;相反地,它會不斷迭代與驗證,直到任務徹底完成。
How it works
Ralph 實作「Ralph Wiggum technique」,使用「帽子系統」讓專門化的人格(如 research、debug、review)透過事件協調執行多步驟任務。它加入「回壓門」——自動檢查(如 lint、型別檢查、測試),會拒絕不完整或破損的工作,迫使代理回到迴圈中修正。系統支援多種 AI 後端(包括 Claude Code、Gemini CLI、Copilot CLI),亦可作為 MCP 伺服器整合。
Who it’s for
想要使用 AI 代理自動化複雜軟體工程任務、且不需要持續人工監控的開發者;以及希望透過 Telegram 加入人類在迴路中的指導者。
Highlights
- Multi-Backend Support: Works with various AI coding assistants like Claude Code, Gemini CLI, and Codex.
- Hat System: Uses specialized roles (e.g.,
code-assist,debug,review) to structure task execution. - Backpressure Gates: Ensures quality by rejecting work that fails tests or linting.
- Human-in-the-Loop: Integration with Telegram allows humans to provide proactive guidance or answer agent questions mid-loop.
- Web Dashboard: An alpha-stage dashboard for monitoring and managing orchestration loops.
- MCP Server Mode: Can be run as a Model Context Protocol server for compatible clients.