rp1-dev

Ready Player One

> stop prompting; start shipping

Professional development workflows for AI coding assistants.
21 commands. 18 specialized agents. Single-pass execution.
Works today with Claude Code. Experimental support for OpenCode and Codex CLI.

---

Why rp1?

Single-pass execution — Constitutional prompts encode expert patterns with built-in rules. No iteration loops, no "let me revise that" — tasks complete in one shot.

Codebase awareness — Run knowledge-build once. Your architecture becomes context for every command. No generic advice — everything respects your patterns.

Lean context architecture — Progressive disclosure and subagent delegation keep your main thread focused. Complex work happens in specialized agents, results flow back clean.

Validate before you build — Hypothesis testing catches bad assumptions early. Design decisions get validated against your codebase before implementation begins.

Start anywhere — Full blueprints with charters and PRDs, or jump straight in with a vague idea. Structured when you need it, flexible when you don't.

---

Quick Start

1. Install the CLI

# macOS / Linux (Homebrew)
brew install rp1-run/tap/rp1

# Windows (Scoop)
scoop bucket add rp1 https://github.com/rp1-run/scoop-bucket
scoop install rp1

# Or use the install script
curl -fsSL https://rp1.run/install.sh | sh

2. Initialize your project

cd your-project
rp1 init

This will set up rp1 and install plugins for your AI assistant (Claude Code, OpenCode, or Codex CLI).

3. Restart your AI tool and run

/knowledge-build    # Generate your codebase knowledge base

Full installation guide →

---

What Can You Do?

Ship a feature

/build user-auth

Review a PR

/pr-review "feature/auth"

Investigate a bug

/code-investigate "bug-123" "Login fails intermittently"

Quick code check

/code-check

See all 21 commands →

---

Principles

Batteries Included	Skills, subagents, and finely-tuned prompts ship out of the box. No assembly required.
Always Open Source	Fully pluggable into existing agentic tools. Your workflows, your control.
Visual-First	Heavily leans on visual language — diagrams, charts, and structured outputs for clarity.
Continuous Evolution	Keep improving and adapting as frontier models mature. Today's best, tomorrow's baseline.
Model/Tool Agnostic	No lock-in to any frontier lab or platform. Works with Claude Code, OpenCode, Codex CLI, and more coming.

---

Platform Support

Platform	Status
Claude Code	Available
OpenCode	Experimental
Codex CLI	Experimental
Cursor, Goose, Amp	Coming Soon

---

Guides

• Feature Development — End-to-end workflow from requirements to verified implementation
• Bug Investigation — Systematic root cause analysis with evidence-based hypothesis testing
• PR Review — Thorough multi-pass analysis with visual diagrams
• Team Onboarding — Get new developers productive on your codebase fast

---

Contributing

See the documentation for architecture and contribution guides. Issues and PRs welcome!

License

Apache 2.0 — See LICENSE

---

Documentation · Issues · Releases