Smart Ralph

Smart Ralph automates spec generation and task execution for developers using Claude Code and Codex. It transforms vague feature ideas into structured research, requirements, design, and actionable tasks, functioning as an AI product team to accelerate development.

How It Works

The system follows a spec-driven workflow with distinct phases: Research, Requirements, Design, and Task Execution, each managed by specialized sub-agents. A key feature is codebase indexing (/ralph-specum:index), which scans existing projects to make components discoverable during research, providing essential context. Execution proceeds task-by-task with fresh context for focused implementation.

Quick Start & Requirements

• Installation:
  • Claude Code: /plugin marketplace add tzachbon/smart-ralph, then /plugin install ralph-specum@smart-ralph.
  • Codex: Use the skill-installer script (e.g., python3 "$CODEX_HOME/skills/.system/skill-installer/scripts/install-skill-from-github.py" --repo tzachbon/smart-ralph --path platforms/codex/skills/ralph-specum).
  • Direct install: /plugin install https://github.com/tzachbon/smart-ralph.
• Prerequisites: Targets Claude Code and Codex. Core loop is self-contained; Codex installation uses Python scripts. No specific hardware (GPU, CUDA) or OS requirements listed.
• Links: Claude Code Plugin Marketplace, GitHub Repo.

Highlighted Details

• Codebase Indexing: Scans project code to generate component specs, improving research context for new features.
• Task-by-Task Execution: Processes tasks sequentially with fresh context for focused implementation.
• Dual Methodologies: Offers ralph-specum (simpler) and ralph-speckit (constitution-based, traceable).
• Approval Gates: Default user approval between phases; --quick option for autonomous execution.

Maintenance & Community

Pull requests are welcomed. The README does not provide explicit community channels (like Discord/Slack), a roadmap, or details on notable contributors or sponsorships.

Licensing & Compatibility

MIT License, generally permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

Failed tasks require manual intervention and resumption via the implement command, with errors logged in .progress.md. Version 3.0.0+ is self-contained, removing the ralph-loop dependency required in v2.x, indicating potential migration considerations.