Every coding agent,
on the same rules.

One home for the rules, skills, and memory your coding agents run on. Write and edit them in onerule, and it syncs every change into Claude Code, Cursor, Codex, and Gemini the moment you save.

See how it works

Free while in beta

Global

Applies to every agent, synced to Claude · Cursor · Codex · Gemini

The baseline every agent reads first. Keep it short. This is the note you'd leave for whoever picks up the work next.

Voice

Plain sentences. No hedging, no filler. If a change is risky, say so before anything else.

Working style

Read before you write. Match the code that's already there.
Make the smallest change that solves the problem.
When you're unsure between two approaches, show both and ask.

Conventions

TypeScript, strict on. Tests sit next to source. Format on save.

pnpm test # vitest pnpm lint --fix pnpm typecheck

Write it once. onerule writes it to every agent.

Each tool reads its own file in its own place. onerule keeps all four identical, so one edit here lands everywhere the moment you save.

Claude Codesynced

~/.claude/CLAUDE.md

Cursorsynced

~/.cursor/rules/onerule.mdc

Codexsynced

~/.codex/AGENTS.md

Geminisynced

~/.gemini/GEMINI.md

Structure your instructions, not just store them.

Skills, shared everywhere

Write a skill once, like commit messages or code review, and every agent inherits it.

Global, then per-agent

One baseline for all of them. Layer agent-specific rules on top without repeating yourself.

Memory that persists

Keep the standing context your agents should never forget. Edit it once and it travels with every tool.

Edit with AI

Describe the change in plain language. onerule rewrites the instruction and syncs it on save.

Every coding agent,on the same rules.