@ainova-systems/intelligence
v0.13.0
Published
Build, version and distribute AI agent intelligence across your organization — one CLI, versioned Intelligence Packages, and a sync engine for Claude Code, Cursor, Copilot, Codex, Pi and OpenCode.
Maintainers
Readme
@ainova-systems/intelligence
npm for AI agent intelligence. One CLI that builds, versions and distributes the rules, agents and skills your AI coding tools run on — rendered natively for Claude Code, Cursor, GitHub Copilot, OpenAI Codex, Pi, OpenCode and AGENTS.md.
npm install -g @ainova-systems/intelligence
cd your-project
intelligence init
intelligence package add @ainova-systems/core
intelligence syncThat's the whole setup. Your project gets:
intelligence.yaml # manifest — what the project consumes, where it goes
intelligence.lock # resolved versions (commit it)
intelligence/ # your own rules / agents / skills
.intelligence/ # installed packages + engine content (gitignored)
.claude/ .cursor/ … # generated per-tool output
AGENTS.mdTeammates need one command after cloning; it restores the locked package store before rendering:
intelligence syncCommands
| Command | What it does |
|---|---|
| intelligence init [--preview\|--apply] | Create, convert, restore or align a project |
| intelligence sync [adapter] | Restore locked content if needed, then render all or one enabled adapter |
| intelligence update [@scope/name] [--preview\|--apply] | Plan or apply project and ranged-package updates |
| intelligence upgrade [--next] [--preview\|--apply] | Replace the installed CLI with the newest version on its npm channel |
| intelligence package add\|remove\|list\|search | Manage versioned Intelligence Packages |
| intelligence adapter list\|create\|enable\|disable\|remove | Manage render adapters |
| intelligence status [--check] | Inspect project state and consistency |
| intelligence registry list\|add\|remove | Manage trusted registry repositories |
Any git repo with rules/, agents/ or skills/ at its root is already an Intelligence Package: install it from an explicit source — intelligence package add github:your-org/your-repo — with versions from its x.y.z git tags. Names (@scope/name) resolve only through registries this project explicitly trusts; a monorepo of packages needs one line in such an index.
Every package has the same compact manifest entry: requested version or
ref only. Repository URL/path, resolved tag/ref and commit SHA are recorded
in the committed intelligence.lock.
Requires git (and on Windows the bash it ships with — Git for Windows). Engine and docs: ainova-systems/intelligence.
