ralph-workflow 0.8.22

Autopilot for coding agents — run Claude Code, Codex, OpenCode, Nanocoder, AGY & Pi unattended with verification, recovery, orchestration.

ralph-workflow

Codeberg is primary: https://codeberg.org/RalphWorkflow/Ralph-Workflow GitHub is a read-only mirror: https://github.com/Ralph-Workflow/Ralph-Workflow (verify: repo-exists)

Autopilot for coding agents — run Claude Code, Codex, OpenCode, Nanocoder, AGY & Pi unattended with verification, recovery, orchestration.

This README is the PyPI-facing surface. The canonical product positioning, who-it-is-for list, install steps, first-run path, trust and safety boundaries, and finish-receipt example all live in the repository root README.md. That page is the single source of truth; PyPI readers should follow the link above for the install → init → diagnose → spec → run → review walkthrough.

At a glance

• Python: >=3.12
• License: AGPL-3.0-or-later
• Runtime: local-first; no required cloud account
• Auth: your agent CLIs authenticate as they always do — Ralph Workflow does not store, read, or proxy credentials

Install and prove the install

pipx install ralph-workflow
ralph --version
ralph --diagnose     # optional pre-flight check

Supported agents

Agent	Notes
Claude Code	Anthropic's CLI for Claude (interactive + headless). Canonical reference agent.
Codex	OpenAI's Codex CLI.
OpenCode	Open-source terminal coding agent.
Nanocoder	Local-only TUI coding agent.
Google Anti Gravity (AGY)	Google's Antigravity CLI (agy, v1.0.9+).
Pi	Minimal coding agent. Headless mode is pi --mode json <prompt>.

Where to go next

• Repository root README.md — canonical product positioning and first-run path
• START_HERE.md — guided first run on one real backlog task
• Operator manual (docs/sphinx/index.rst) — maintained reference, configuration, troubleshooting, API reference
• CONTRIBUTING.md — contribution intent and source-of-truth ordering
• Package-side CONTRIBUTING.md — Python-specific dev build, policy-driven pipeline, guardrails
• Issue tracker