ralphex workflow

Autonomous plan execution task by task: each task in a fresh Claude Code or Codex session, then a multi-pass review. Written in Go

  • 1,495 ★ on the repository
  • Last commit
  • MIT
Workflow

Broad access

Why this level

  • Edits the repository autonomously
  • Token spending is hard to predict
  • Unattended runs may follow a different billing scheme
All reasons and checks

umputun/ralphex

Install

Manual install

brew install umputun/apps/ralphex

Or go install github.com/umputun/ralphex/cmd/ralphex@latest or a Docker image. Launch in a separate branch.

Access and checks

How we review

Broad accessThe tool can change data, run commands or handle secrets.

  • Edits the repository autonomously
  • Token spending is hard to predict
  • Unattended runs may follow a different billing scheme

This is third-party code. Review the repository files before installing.

What it does

ralphex executes a plan task by task autonomously: each task runs in a fresh Claude Code or Codex session and then goes through a multi-pass review. Installation: `brew install umputun/apps/ralphex`, `go install github.com/umputun/ralphex/cmd/ralphex@latest` or a Docker image. Version 1.7.0 from 07.09.2026, MIT license. The README warns that unattended `claude -p` runs may be counted against a separate monthly Agent SDK credit pool and not the subscription. Plan terms may change, so check your own plan.

Who it is for. For developers who hand an agent a big plan for many hours and want a multi-pass review of each task.

Good fit when

  • You need to execute a plan of many tasks without your involvement
  • You need a fresh session for each task
  • You need a multi-pass review of the result

Not a fit when

  • You cannot let an agent edit a repository autonomously
  • You cannot spend tokens unattended

Example request

Execute the plan from the file task by task in a separate branch, review after each task and show the final diff

Limitations

The utility edits the repository autonomously and launches agents many times, so access is wide and token spending is hard to predict. The README warns about a possible separate billing scheme for unattended runs: check your plan terms. Work in a separate branch and in a container.

How to disable. Remove the skill from the agent's folder or the plugin through the plugin manager.

FAQ

What does ralphex do?

It executes a plan task by task autonomously, each task in a fresh session, then a multi-pass review.

How do I install it?

brew install umputun/apps/ralphex, go install or a Docker image.

What about costs?

The README warns that unattended runs may be counted in a separate pool: check your plan.

Discussion

No questions or reviews yet.

Workflows

A CLI orchestrator that loops Claude Code, Codex, Gemini CLI and other agents until a task is done and blocks incomplete work with tests and lint

CLIExtended access3.2KRepository stars
Needs a VPN

Skills for cross-provider work between Claude Code and Codex: one agent drafts the plan, the other reviews it, then a build and an independent code inspection f

SkillExtended access2.9KRepository stars

oh-my-codex

oh-my-codex (OMX)

Needs a VPN

A layer on top of Codex CLI: one flow from clarifying a task to verification, agent teams, hooks, a HUD and state in .omx/

CLIBroad access33.5KRepository stars
Editors’ pick

Multi-agent orchestration for Claude Code: teams, roles, autopilot and task routing across skills and subagents

PluginBroad access39.8KRepository stars

Plugin marketplace and CLI for Claude Code: agent swarms, background workers, vector memory, security review and test generation

PluginBroad access74.4KRepository stars

A terminal app for running several Claude Code, Codex, Gemini CLI and Aider sessions in parallel, each in its own tmux session and git worktree

CLIBroad access8.6KRepository stars

Comparison with similar tools

Entries that share the main search query, and how they differ.

EntryOfficialRepository starsMain difference
ralphex (this entry)no1,495
Continuous Claudeno1,382a Claude Code loop with pull requests, checks and merge
TAKTno1,403orchestration of several agents by YAML workflows
Spec Kittyno1,681spec-driven development, version 4 is still a release candidate

README in short

The README describes autonomous plan execution task by task in fresh Claude Code or Codex sessions with a multi-pass review, installation through Homebrew, go install and Docker. MIT license.

README badge

Report a problem

Foxx AIralphex
Foxx AI

I am Foxx AI and I have already vetted this tool. Ask about install, setup or anything else, and I will keep it simple.