Managed AI Dev Practice

Your AI agents
need a harness.

Rules, hooks, workflows, and shared team memory that make your AI coding agents reliable. Works with Claude Code, Cursor, and Copilot.

Get started
The problem

AI agents are powerful. Without structure, they're chaos.

“3 hours configuring CLAUDE.md. The code still doesn't lint.”
Prompts are advisory. Agents follow them ~80% of the time. Only hooks give 100%.
“My agent ignored half the rules. Again.”
Hand-written rules lack structure. Agents need tested, research-backed conventions — not vibes.
“Set it up once. Now it's outdated and nobody updates it.”
Claude ships updates weekly. Cursor evolves constantly. Your rules rot unless someone maintains them.
“Each dev's agent does something different. Same repo.”
Without shared conventions, every agent works differently. No consistency, no shared context, no team memory.
What you get

Your agents. Our structure.

RIG doesn't replace your AI tools. It adds the practice layer they're missing — so the code they write actually ships.

✓

Deterministic enforcement

Pre-commit hooks catch what prompts miss. Lint, typecheck, security gates before every commit. Hooks = 100%. Prompts ≈ 80%.

■

Opinionated defaults

Code review, security rules, agent orchestration — pre-configured from real research. Not opinions. Not vibes.

↻

Living updates

New research, better prompts, improved workflows. rig update pulls the latest. Your customizations stay untouched.

“The more autonomy you give AI agents,
the more structure they need. RIG is that structure.”
1.7x more bugsAI-generated code vs human-written. CodeRabbit, 2025
≈80% complianceCLAUDE.md rules are advisory. Only hooks give 100%
32K token dropAccuracy degrades past ~32K context. Stanford & UC Berkeley
Team features

One harness. Whole team aligned.

When your team uses RIG, every agent — on every machine — follows the same conventions. No more “works on my machine” for AI.

📋

Shared conventions

Commit CLAUDE.md and .cursor/rules/ to your repo. Every teammate's agent inherits the same rules — code style, review process, security.

🔗

Shared context

Agents share PROGRESS.md and DECISIONS.md — status, blockers, architectural decisions. Your PM's agent knows what the dev's agent just shipped.

🧠

Coming Shared memory

Persistent memory across machines — patterns learned, mistakes avoided. Your team's AI gets smarter with every session.

How it works
01
npx create-rig
Pick your preset and stack. RIG generates CLAUDE.md, agents, rules, hooks, and workflows. 30 seconds.
02
code with AI
Your agents now follow your rules. Pre-commit hooks catch mistakes. Code review runs automatically. You ship faster.
03
rig update
Claude changed? Cursor updated? We ship new rules. One command — your setup stays current. No manual maintenance.

Get started

Leave your email — we'll send setup instructions and early updates.

Free forever. No spam.
FAQ

Common questions

I already have CLAUDE.md. Why do I need RIG?
I already have CLAUDE.md. Why do I need RIG?
What is RIG?
RIG is a managed set of rules, hooks, and workflows for AI-assisted development. Install with npx create-rig, and it generates config files that make your AI agents produce more reliable code. It's a practice layer — not a new tool, but structure for the tools you already use.
What are AI coding agents?
Tools like Claude Code, Cursor, GitHub Copilot that write and edit code for you. They're powerful — but without structure, they ignore conventions, skip tests, and produce inconsistent results. RIG adds that structure.
Does RIG provide its own agents?
No. Your agents come from your LLM provider and your tools. RIG provides the harness — rules they follow, hooks that enforce quality, workflows that keep your codebase healthy. You keep all your agents, plugins, and custom skills.
How does it work technically?
RIG generates CLAUDE.md for Claude Code, .cursor/rules/ for Cursor, AGENTS.md for Copilot. Plus pre-commit hooks. All from one rig.toml config. Sections marked [RIG-MANAGED] get updated by rig update; your customizations stay untouched.
How does it help teams?
Generated files are committed to git — any teammate who clones the repo gets the same conventions, hooks, and workflows instantly. Agents share context through PROGRESS.md and DECISIONS.md. We're building shared persistent memory next.
Is it free?
Yes. The core product is and will remain free. Premium content packs (industry-specific rules, advanced workflows) are planned for the future.
Will my customizations survive updates?
Yes. RIG marks managed sections with [RIG-MANAGED] tags. rig update only touches those sections. Everything you've customized stays untouched.