Intelligence Dev Packs — Development skill for Claude Code

intelligence-dev-packs

Shared AI-first engineering practices, packaged as portable rules, agents, and skills. Author the workflow once, reuse it in every repository, and let both your team and your AI coding agents read the same source.

Built and maintained by [Ainova Systems](https://www.ainovasystems.com). Extracted from running multiple AI-coded enterprise systems in daily production, where AI writes the majority of the code.

Quick start

The packs ship as [Intelligence Packages](https://github.com/ainova-systems/intelligence), installed with the Intelligence CLI. Needs Node.js 18+, git, bash and awk.

npm install -g @ainova-systems/intelligence

cd your-project
intelligence init --targets claude              # claude, cursor, copilot, codex, pi, opencode
intelligence registry add https://github.com/ainova-systems/intelligence-dev-packs.git
intelligence package add @ainova-systems/core   # add @ainova-systems/spec for spec-driven projects

`init` writes `intelligence.yaml` and `intelligence.lock` - both committed - and renders each enabled tool's native files. `registry add` is what makes this repository's package names resolvable: names resolve only through registries the project has explicitly trusted, so nothing installs from a guessed URL. `package add` records the name and pins the resolved tag and commit in the lock.

Then run **`dev-init`** so the agent pins the project profile, creates the `ai:*` labels, copies the pack PR template when the repo has none, and merges the harness permission rules. It reports what it filled, what it left, and any project rule that overlaps a package rule. It does not commit. If the spec pack is installed, `spec-init` is the next step.

What you get

Pack	Install when	Contents
core (dev-, git-) @ainova-systems/core	Always - it is universal	6 always-on rules (skill-first, context engineering, verification gates, rollback safety, commit conventions, git workflow),